ARTICLE IN PRESS
Behaviour Research and Therapy 46 (2008) 836– 844

Contents lists available at ScienceDirect

Behaviour Research and Therapy journal homepage: www.elsevier.com/locate/brat

Prevalence and correlates of hoarding behavior in a community-based sample
Jack F. Samuels a,Ã, O. Joseph Bienvenu a, Marco A. Grados a, Bernadette Cullen a, Mark A. Riddle a, Kung-yee Liang b, William W. Eaton c, Gerald Nestadt a a Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, 600 N. Wolfe Street, Meyer 109, Baltimore, MD 21287-7228, USA Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21287, USA c Department of Mental Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21287, USA b a r t i c l e in fo
Article history: Received 14 February 2008 Received in revised form 7 April 2008 Accepted 8 April 2008 Keywords: Hoarding Prevalence Risk factors Personality disorders Comorbidity

abstract
Little is known about the prevalence and correlates of hoarding behavior in the community. We estimated the prevalence and evaluated correlates of hoarding in 742 participants in the Hopkins Epidemiology of Personality Disorder Study. The prevalence of hoarding was nearly 4% (5.3%, weighted) and was greater in older than younger age groups, greater in men than women, and inversely related to household income. Hoarding was associated with alcohol dependence; paranoid, schizotypal, avoidant, and obsessive–compulsive personality disorder traits; insecurity from home break-ins and excessive physical discipline before 16 years of age; and parental psychopathology. These findings suggest that hoarding may be relatively prevalent and that alcohol dependence, personality disorder traits, and specific childhood adversities are associated with hoarding in the community. & 2008 Elsevier Ltd. All rights reserved.

Introduction Hoarding behavior has been called ‘pathological collecting’ and is characterized by the acquisition of, and unwillingness or inability to discard, large quantities of seemingly useless objects (Frost & Gross, 1993; Greenberg, Witzum, & Levy, 1990). The behavior can lead to significantly cluttered living space in the home and can cause considerable distress and impairment in functioning for individuals and their family members (Tolin, Frost, Steketee, & Fitch, 2008). The clutter may interfere with the normal use of space for basic household activities and increase the risk of injuries due to fire and falling, and illnesses due to poor sanitation (Frost, Steketee, & Williams, 2000; Steketee, Frost, & Kim, 2001). To date, hoarding behavior has been studied primarily in samples of individuals who were not selected from the community, such as respondents to advertisements for hoarding research studies, participants in hoarding self-help groups, individuals referred to clinics specializing in the treatment of obsessive–compulsive disorder, and individuals participating in family studies of obsessive–compulsive disorder. Results from these studies suggest that individuals with hoarding behavior have more symptoms of anxiety and depression, a greater prevalence of anxiety disorders, especially generalized anxiety disorder and social phobia, and poorer functioning (Frost, Steketee, Williams, & Warren 2000; Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002). Moreover, individuals with hoarding behavior are more likely to exhibit

à Corresponding author. Tel.: +1 410 614 4942; fax: +1 410 614 8137.

E-mail address: jacks@jhmi.edu (J.F. Samuels). 0005-7967/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.brat.2008.04.004

ARTICLE IN PRESS
J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844 837

obsessive–compulsive, dependent, avoidant, and schizotypal personality traits (Samuels, Bienvenu, et al., 2007; Samuels, Shugart, et al.,2007). However, little is known about the prevalence, and sociodemographic and clinical correlates, of hoarding behavior in community samples. Given that compulsive hoarding occurs in approximately 30% of individuals with obsessive– compulsive disorder (OCD) in clinical and family samples (Rasmussen & Eisen, 1992; Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002), and that the population prevalence of OCD is estimated to be about 1–2% (Karno, Golding, Sorenson, & Burnam, 1988), based on population-based surveys, the population prevalence of compulsive hoarding behavior has been estimated to be approximately 0.4%; however, since hoarding behavior can occur in individuals without OCD, this may be an underestimate (Steketee & Frost, 2003). Furthermore, knowledge about the demographic and clinical characteristics of individuals with hoarding behavior is restricted to these particular samples of individuals, who may not reflect the wider range of hoarding behavior in the community. A biological basis for the development of hoarding behavior has been suggested from case reports of hoarding emerging after traumatic brain lesions (Anderson, Damasio, & Damasio, 2005), and from functional imaging studies identifying differential activation of specific brain regions in OCD patients with and without hoarding behavior (Saxena et al., 2004). A genetic etiology for hoarding behavior is supported by the occurrence of hoarding behavior in known genetic syndromes (e.g., Prader-Willi syndrome and velocardiofacial syndrome) (Dykens & Shah, 2003; Gothelf et al., 2004), aggregation of hoarding behavior in OCD families (Samuels, Bienvenu, et al., 2007), and genetic linkage of hoarding behavior to specific chromosomal regions in families with OCD (Samuels, Shugart, et al., 2007). However, in most cases, precipitating factors for the development of hoarding behavior have not been identified. Although early material deprivation has been hypothesized as a risk factor for hoarding behavior (Steketee & Frost, 2007), this was not supported by the only study to evaluate it empirically (Frost & Gross, 1993). More recently, two studies found that traumatic life events are associated with hoarding behavior. Hartl, Duggany, Allen, Steketee, and Frost (2005) found that, compared with controls, participants with hoarding reported having experienced greater frequency and greater number of different types of traumatic events, especially having had something taken by force, being physically handled roughly in childhood or adulthood, and being forced to engage in sexual activity in childhood or adulthood. Cromer, Schmidt, and Murphy (2007) reported that, in patients admitted to an adult OCD clinic, those with hoarding reported significantly more traumatic life events, and that the number of traumatic events experienced was significantly related to hoarding severity, even when controlling for current age, age at onset of OCD, and symptoms of depression and anxiety. Little is known about gender-specific differences between individuals with and without hoarding behavior. Wheaton, Cromer, LaSalle-Ricci, and Murphy (2008) reported that, among female patients in an OCD clinic, those with hoarding behavior had earlier age at onset of OCD, greater severity of OCD symptoms, and greater prevalence of bipolar disorder, panic disorder, binge-eating disorder, and alcohol and substance use disorders; in contrast, among men, the only difference was a greater prevalence of social phobia in those who hoard. To our knowledge, no studies have evaluated correlates of hoarding behavior separately in men and women in a community sample. In the current study, we investigated hoarding behavior in a community sample of adults who participated in an epidemiologic study of personality and personality disorders. The aims of the study were: (1) to estimate the prevalence of hoarding, overall and by sociodemographic characteristics, in this community sample; (2) to investigate the association between hoarding behavior and potential clinical correlates (personality disorder and personality dimensions; history of psychiatric disorders; and current functioning); (3) to investigate the association between hoarding behavior and specific self-reported childhood adversities, including parental psychopathology and specific childhood traumas; and (4) to determine if the relationships between hoarding and specific correlates are different in men and women.

Method Participants As described previously (Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002), participants in the Hopkins Epidemiology of Personality Disorder Study were sampled from the Baltimore Epidemiologic Catchment Area (ECA) Followup survey (Eaton et al., 1997). In brief, 3481 adult household residents of east Baltimore were sampled probabilistically and were interviewed in 1981–1982, using the Diagnostic Interview Schedule (DIS); 810 of these individuals also were examined by psychiatrists as part of the Clinical Reappraisal (Anthony et al., 1985). Between 1993 and 1996, 1920 (73%) of the surviving participants were re-interviewed. From these 1920 individuals, all those who were examined by psychiatrists in 1981, as well as all those who were identified by the DIS as having a lifetime diagnosis of mania, depression, panic disorder, OCD, alcohol use disorders, or drug use disorders at follow-up, were selected. In addition, a 25% random sample was selected from the remaining participants. Of the 1258 individuals selected with these criteria, 516 could not be interviewed because they could not be traced; refused participation; were deceased; or were too ill to participate. A total of 742 individuals completed the personality examinations between 1997 and 1999. The gender and ethnic distributions of these participants were similar to those of the 516 individuals who were not interviewed; however, the interviewed individuals were younger, on average, than the non-interviewed. The gender and ethnic distributions of the study

ARTICLE IN PRESS
838 J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844

participants also were similar to those of the 3481 individuals examined in 1981, although the study participants were younger (Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002). Materials As described previously (Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002), personality disorder criteria were assessed by psychologists using the International Personality Disorder Examination (IPDE) (Loranger et al., 1994), a semi-structured instrument designed to be administered by clinicians to detect all the relevant criteria for diagnosis of all DSM-IV personality disorders (American Psychiatric Association, 1994). The psychologists were directed to evaluate abnormal personality traits manifest over the subject’s entire adult life. Each criterion was rated ‘0’ (absent), ‘1’ (accentuated or exaggerated), ‘2’ (criterion level or pathological), or ‘9’ (missing or unknown). A dimensional score was calculated for each of the 10 DSM-IV personality disorders by counting the number of traits of the specific disorder that were rated ‘2.’ The hoarding criterion for obsessive–compulsive personality disorder, as specified in DSM-IV, is ‘‘unable to discard worn-out or worthless objects even when they have no sentimental value’’ (American Psychiatric Association, 1994). To address this criterion, the interviewers were to ask participants the following questions: ‘‘Do you find it almost impossible to throw out worn-out or worthless things? If so, is that true even when they don’t have any sentimental value? Give me some examples. Is this a problem for you or for others? If so, tell me about it.’’ The interviewers also were encouraged to cross-examine the participants in order to establish the presence of the behavior. As for other personality traits, the psychologists were directed to evaluate the hoarding criterion as manifest over the subject’s entire adult life, and to rate the criterion as ‘0’ (absent), ‘1’ (accentuated or exaggerated), ‘2’ (criterion level or pathological), or ‘9’ (missing or unknown). For this paper, a rating of ‘2’ was considered to indicate clinically significant hoarding behavior. Potentially traumatic childhood experiences were assessed by the psychologists, with questions about parental death, parental separation or divorce, conflict between parents, not being raised by both parents, inappropriate sexual contact with a family member, and lack of security from break-ins, all before the age of 16 years. Information about the presence of depression, mania, and heavy drinking in natural mother and natural father was obtained by lay interviewers in the original ECA Study in 1982 and was used in the present analyses. Psychiatric diagnoses were made by psychiatrists who examined participants using the Schedule for Clinical Assessment in Neuropsychiatry (SCAN, version 1.5) (Wing et al., 1990) for current and lifetime disorders according to DSMIIIR criteria (American Psychiatric Association, 1987). The probes in the OCD section specifically included questions about checking compulsions, ordering compulsions, and cleaning compulsions, as well as a general probe for obsessions, but did not include questions about hoarding symptoms. The psychiatrists did not have access to the results of the psychologist’s assessment of personality before conducting their examinations. Global functioning was scored by the psychiatrists, using the Global Assessment of Functioning scale (GAF). Severity of stressors in nine domains (family; occupational; financial; legal; present living situation; marital; parental; social; and physical) also was rated by the psychiatrists; the scale for each domain ranged from 1 (‘none’) to 7 (‘catastrophic’) (American Psychological Association, 1994). A total stressor scale, which summed the scores across all stressor domains, was used in the current analyses. Procedure The study protocol was approved by the Institutional Review Board of the Johns Hopkins Medical Institutions. The assessment of personality disorder criteria and childhood experiences was conducted in person by four masters-level clinical psychologists, who obtained informed consent from participants prior to beginning the interview. Following each participant interview, the psychologist interviewed an informant about the participant, using questions from the IPDE and scored as above; formulated a final rating for each criterion based on her clinical judgment of both participant and informant reports; and completed a case summary describing the personality of each participant. On a separate visit, the psychiatric examinations were conducted in person by one of five Johns Hopkins School of Medicine psychiatrists trained in use of the SCAN. Data analysis The unweighted prevalence of hoarding across sociodemographic characteristics was compared using the w2 test. The odds ratio (OR) and 95% confidence interval (CI) of the association between each potential correlate and hoarding were estimated using logistic regression. Logistic regression also was used to evaluate the association between hoarding and potential correlates, controlling for other correlates in the model. We used separate models to evaluate the magnitude of these relationships in men and women, respectively. We also evaluated the interactions between the clinical correlates and gender; in each of these models, hoarding was the dependent variable; one independent variable was the specific clinical characteristic, and the other independent variable was sex; and a term for the interaction between sex and the clinical characteristic was included.

ARTICLE IN PRESS
J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844 839

To estimate the weighted prevalence of the hoarding in the eastern Baltimore population, we used weights to account for the unequal selection probabilities. The weights are the product of the two selection probabilities from the baseline study and the subsequent screening at follow-up, as described above (Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002). Weighted results better represent the population, whereas unweighted results better represent the sample.

Results Prevalence and description of hoarding behavior Of 735 participants with information on the hoarding trait, 27 (3.7%) were rated as having ‘pathological’ hoarding. The weighted prevalence of hoarding was 5.3%. The individuals with hoarding described substantial difficulties due to this behavior. For example, one participant, a 49-year-old man, said that ‘‘My room is like a bomb hit it. I’ve got books and papers, stuff in the corner there. I don’t want to throw nothing away. Old suits in my closet, I know I’ll never wear again in my life. Old beat up tennis shoes, think I’ll find a use for them. I never throw a book away. I like to keep articles, the whole paper; it starts building up on me in a hurry. Newspapers knee-high. I keep a whole drawer full of rubber bands; don’t know why, but I do. Lots of junk.’’ Another participant, a 41-year-old woman, noted that she has saved ‘‘old clothes from the 1970s, piled up clocks, iron, tiny ´ television, picture frame, fans. My house sometimes looks like a junk shop. I argue with my fiance over throwing things away; he wants to get rid of all my good stuff; to me it’s good stuff, to him it’s junk. Stuff I’ve had for years, reminds me of my motheryyou know I’m not going to throw that away.’’
Table 1 Relationship between hoarding and sociodemographic characteristics, Eastern Baltimore, 1997– 1999 Prevalence N (%) Age (in years) 34–44 (N ¼ 266) 45–54 (N ¼ 244) 55–94 (N ¼ 225) Sex Women (N ¼ 465) Men (N ¼ 270) Race/ethnicity White (N ¼ 442) Other (N ¼ 293) Current marital status Married/cohab (N ¼ 374) Sep/Divorced (N ¼ 180) Never married (N ¼ 99) Widowed (N ¼ 82) Lives alone No (N ¼ 594) Yes (N ¼ 141) Highest grade completed Post-HS (N ¼ 270) HS graduate (N ¼ 213) Not HS graduate (N ¼ 247) Currently employed Yes (N ¼ 470) No (N ¼ 262) Household income, annual $ 449,999 (N ¼ 174) 20,000–49,999 (N ¼ 246) o20,000 (N ¼ 221) a b

Odds ratio (95% CI)

6 (2.3) 7 (2.9) 14 (6.2)

1.00 1.28 (0.4–3.9) 2.88 (1.1–7.6)a

12 (2.6) 15 (5.6)

1.0 2.22 (1.02–4.8)a

14 (3.2) 13 (4.4)

1.0 1.42 (0.7–3.1)

11 6 5 5

(2.9) (3.3) (5.1) (6.1)

1.00 1.14 (0.4–3.1) 1.76 (0.6–5.2) 2.14 (0.7–6.3)

20 (3.4) 7 (5.0)

1.00 1.50 (0.6–3.6)

10 (3.7) 7 (3.3) 10 (4.0)

1.00 0.88 (0.3–2.4) 1.10 (0.4–2.7)

13 (2.8) 14 (5.3)

1.00 1.98 (0.92–4.3)

2 (1.1) 8 (3.3) 11 (5.0)

1.00 2.89 (0.6–13.8) 4.51 (0.99–20.6)b

po0.05. p ¼ 0.052.

ARTICLE IN PRESS
840 J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844

As shown in Table 1, the prevalence of hoarding increased with age, from 2.3% in the youngest, to 6.2% in the oldest, age groups; the odds of hoarding was nearly three times as great in the oldest compared with the youngest age group. In addition, the prevalence of hoarding was over two times as great in men (5.6%) compared with women (2.6%). The prevalence of hoarding was about two times as great in those who were widowed compared with those who were currently married, and two times as great in the currently unemployed compared with the employed, although the differences were not statistically significant at po0.05. The prevalence of hoarding was not substantially different by education, living arrangement (i.e., lives alone or with others), or race/ethnicity. The prevalence of hoarding was inversely related to household income; the odds of hoarding was over four times as great in the poorest, compared with the wealthiest, households (p ¼ 0.052). Household income was strongly related to age, sex, race, marital status, living arrangement, education, and employment status (a greater proportion of participants in lower-income households were older, female, non-whites, not currently married, living alone, unemployed, and of lower educational attainment). However, controlling for these variables, one by one, in logistic regression models did not appreciably change the magnitude of the association between hoarding and household income (results not shown). Association of hoarding with clinical characteristics The current total psychosocial stressor score was similar in participants with (M ¼ 13.6, SD ¼ 3.8) and without (M ¼ 13.9, SD ¼ 4.1) hoarding; t(567) ¼ 0.31, p ¼ 0.76. However, the current GAF score was significantly lower in those with hoarding (M ¼ 67.9, SD ¼ 15.3) than in those without (M ¼ 76.1, SD ¼ 10.0); t(700) ¼ 4.2, po0.001). The prevalence of almost all lifetime, as well as current, Axis I disorders was not significantly different in participants with and without hoarding (results not shown). However, the lifetime prevalence of alcohol dependence was significantly greater in individuals with (52.2%), compared with those without (19.5%), hoarding behavior (OR ¼ 4.5, 95% CI ¼ 1.9–10.4, po0.001). Current alcohol dependence was more prevalent in individuals with hoarding (11.1%) than in those without hoarding (4.8%), although the difference was not significant (OR ¼ 2.5, 95% CI ¼ 0.7–8.7, p ¼ 0.20). None of the individuals with hoarding behavior was diagnosed with OCD; however, some participants with hoarding might have been diagnosed with OCD had the SCAN probe questions included an assessment of compulsive hoarding. The odds of hoarding increased with the number of personality disorder traits, including paranoid (OR ¼ 1.60, per unit increase in number of paranoid traits), schizotypal (OR ¼ 1.49), antisocial (OR ¼ 1.2), avoidant (OR ¼ 1.66), and obsessive–compulsive traits (excluding the hoarding trait) (OR ¼ 1.76) (Table 2). Adjusting for lifetime alcohol dependence substantially reduced the magnitude of the association between hoarding and antisocial personality disorder (OR ¼ 1.07, 95% CI ¼ 0.8–1.4), but not the other personality disorder dimensions. Association of hoarding with childhood adversities As shown in Table 3, several childhood adversities were reported significantly more often by participants with hoarding. The odds of hoarding were nearly three times as great in individuals reporting having a parent with psychiatric symptoms (depression, mania, or heavy drinking). Moreover, the odds of hoarding were nearly four times as great in participants reporting lack of security from home break-ins in childhood. In addition, the odds of hoarding were over four times as great in individuals who reported receiving excessive physical discipline in childhood. We evaluated the relationships between these four child adversities and hoarding in a series of logistic regression models, controlling one by one for each of the sociodemographic characteristics (age, sex, household income), personality disorder dimensions (paranoid, schizotypal, avoidant, and obsessive–compulsive), and other clinical features (GAF score, lifetime alcohol dependence) that were associated with hoarding in previous analyses. We found that, in general, the
Table 2 Relationship between hoarding and number of personality disorder traits, Eastern Baltimore, 1997– 1999 Odds ratio (95% CI) Paranoid Schizoid Schizotypal Antisocial Borderline Histrionic Narcisssistic Avoidant Dependent Obsessive– compulsived po0.01. po0.001. po0.05. d Excludes the trait ‘‘unable to discard worn-out or useless objects’’. b c a

1.60 1.53 1.49 1.24 1.28 1.01 1.24 1.66 1.49 1.76

(1.1– 2.2)a (1.0– 2.3) (1.2– 1.9)b (1.02– 1.5)c (0.98– 1.7) (0.6– 1.8) (0.8– 1.9) (1.2– 2.2)b (0.9– 2.3) (1.3– 2.3)b

ARTICLE IN PRESS
J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844 841

Table 3 Relationship between hoarding and childhood adversities, Eastern Baltimore, 1997– 1999 Hoarding N (%) Conflict between parents Family sexual contact Raised by both parents Death of mother Parents separated or divorced Death of father Death of mother or father Psychiatric symptoms, fathera Psychiatric symptoms, mothera Lack of security from home break-ins Excessive physical discipline a b c

Nonhoarding N (%) 229 63 217 32 150 10 95 151 111 30 144 (33.1) (8.9) (31.7) (4.6) (23.0) (10.3) (14.4) (25.4) (18.0) (4.3) (20.5)

Odds ratio (95% CI) 1.14 1.27 1.54 1.71 2.09 2.28 2.31 2.69 2.73 3.90 4.18 (0.5–2.6) (0.4–4.3) (0.7–3.5) (0.4–7.6) (0.9–4.7) (0.8–6.3) (0.9–5.7) (1.2–6.2)b (1.2–6.4)b (1.3–12.0)b (1.9–9.1)c

9 3 10 2 10 5 7 11 9 4 14

(36.0) (11.1) (41.7) (7.7) (38.5) (20.8) (28.0) (47.8) (37.5) (14.8) (51.9)

Depression, mania, or heavy drinking. po0.05. po0.001.

Table 4 Relationship between hoarding and childhood adversities, controlling individually for sociodemographic and clinical characteristics, Eastern Baltimore, 1997– 1999 Psychiatric symptoms, father Controlling for: – Age Sex Household income Number of paranoid traits Number of schizotypal traits Number of avoidant traits Number of obsessive– compulsive traitsa GAF score Alcohol dependence, lifetime Psychiatric symptoms, father Psychiatric symptoms, mother Home break-ins Excessive physical discipline a Psychiatric symptoms, mother

Home breakins

Excessive physical discipline

2.7 2.6 2.9 2.8 2.6 2.7 3.0 2.4 2.2 1.9 – 2.5 2.5 2.3

(1.2– 6.2) (1.1– 6.1) (1.3– 6.8) (1.1– 7.4) (1.1– 6.2) (1.1– 6.3) (1.2– 7.2) (1.03– 5.8) (0.9– 5.6) (0.7– 4.8) (1.1– 5.8) (1.1– 5.9) (0.99– 5.5)

2.7 3.1 3.1 3.3 2.6 2.4 2.5 2.8 1.9 2.9 2.2 – 2.6 2.3

(1.2– 6.4) (1.3– 7.3) (1.3– 7.3) (1.3– 8.4) (1.1– 6.3) (1.01– 5.8) (1.01– 6.1) (1.2– 6.6) (0.7– 5.1) (1.1– 7.6) (0.9– 5.5) (1.1– 6.2) (0.97– 5.5)

Odds ratio (95% CI) 3.9 (1.3– 12.0) 5.1 (1.6– 16.2) 4.6 (1.4– 14.3) 5.6 (1.7– 18.7) 3.8 (1.2– 11.9) 3.3 (1.04– 10.4) 3.3 (1.03– 10.5) 3.1 (0.97– 10.2) 2.8 (0.8– 9.1) 4.6 (1.4– 14.8) 2.6 (0.7– 9.6) 4.1 (1.3– 12.9) – 3.5 (1.1– 11.2)

4.2 4.2 4.0 4.0 4.0 3.5 4.1 4.1 3.3 3.8 4.2 4.6 4.0 –

(1.9– 9.1) (1.9– 9.1) (1.8– 8.8) (1.7– 9.8) (1.8– 9.1) (1.6– 7.8) (1.8– 9.1) (1.9– 9.1) (1.4– 7.6) (1.6– 8.8) (1.8– 9.9) (2.0– 10.6) (1.8– 8.8)

Excludes the trait, ‘‘unable to discard worn-out or useless objects’’.

magnitude of the relationships between the four child adversities and hoarding did not substantially change after these adjustments. Moreover, each child adversity remained independently associated with hoarding when each of the other three adversities was included one by one in the models (Table 4). Gender interactions We examined the relationship between childhood adversities, personality disorder dimensions, and other clinical characteristics, on the one hand, and hoarding, on the other, separately in men and women. We found that the magnitude of the association of hoarding with paternal psychiatric symptoms was considerably stronger in women (OR ¼ 6.97, 95% CI ¼ 1.8–26.8, po0.001) than in men (OR ¼ 1.23, 95% CI ¼ 0.3–4.7); in a logistic model, the interaction term b ¼ 1.74, SE ¼ 0.97, p ¼ 0.07). Moreover, the magnitude of the association of hoarding with maternal psychiatric symptoms was considerably stronger in women (OR ¼ 4.68, 95% CI ¼ 1.4–15.7, po0.01) than in men (OR ¼ 1.91, 95% CI ¼ 0.5–7.4); interaction term b ¼ 0.90, SE ¼ 0.93, p ¼ 0.33). In addition, the relationship with avoidant personality disorder score was substantially stronger in women (OR ¼ 2.45, 95% CI ¼ 1.6–3.7; po0.001) than in men (OR ¼ 1.09, 95% CI ¼ 0.6–1.9); interaction term b ¼ 0.81, SE ¼ 0.36, p ¼ 0.03). The magnitude of the association between hoarding and lifetime alcohol dependence was similar in women (OR ¼ 3.91, 95% CI ¼ 0.91–16.8) and men (OR ¼ 3.37, 95% CI ¼ 1.2–9.8). The magnitude of associations between hoarding and other correlates also were similar in women and men (results not shown). Discussion There are four major findings from this study of hoarding behavior. First, the prevalence of hoarding behavior in this community sample was nearly 4% (5%, weighted to the population). This is considerably higher than estimates of about

ARTICLE IN PRESS
842 J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844

0.4% in the community, based on the known population prevalence of OCD and the proportion of OCD cases with compulsive hoarding (Steketee & Frost, 2003). This estimate is based on the notion that most, if not all, individuals with hoarding behavior have OCD. However, currently there is debate about the classification of hoarding, and whether the behavior characterizes a subtype of OCD, or a unique syndrome, is an unresolved clinical issue (Steketee & Frost, 2003). On the one hand, it has been reported that approximately one-third of individuals with OCD have hoarding symptoms (Rasmussen & Eisen, 1992; Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002), and hoarding obsessions and compulsions are considered as symptoms of OCD in the Yale–Brown Obsessive–Compulsive Scale—Symptom Checklist (Goodman et al., 1989). On the other hand, it has been reported that hoarding correlates poorly with other symptoms of OCD (Wu & Watson, 2005), and that patients with compulsive hoarding have different clinical profiles, response to treatment, and functional neuroimaging findings compared with other OCD patients (Saxena, 2007). The results of the current study would suggest that, in this community sample, hoarding behavior can occur without OCD; indeed, none of the hoarding individuals was diagnosed with OCD by examining psychiatrists. However, some of these individuals probably would have been diagnosed with OCD, if the SCAN probes had included a question about compulsive hoarding. Based on the phenomena recorded during the IPDE interviews, we consider it likely that at least some of the participants with hoarding would meet criteria for OCD (with compulsive hoarding). Moreover, although none of the 13 participants with psychiatrist-diagnosed OCD received a rating of ‘pathological’ on the hoarding item, 4 (31%) were rated as ‘accentuated or exaggerated,’ indicating evidence for sub-threshold hoarding behavior in these individuals. Second, the prevalence of hoarding differed by demographic characteristics. Consistent with impressions from prior studies of individuals with hoarding, we found that the prevalence of hoarding behavior increased markedly with age (Steketee & Frost, 2003), and was nearly three times as prevalent in the oldest compared with the youngest age group. We do not know if this reflects an actual increase in the incidence of hoarding, or its severity, because we did not assess the age at onset of the behavior. It may be that the consequences of a hoarding tendency become more problematic as individuals age, due to accretion of objects over time, or because of increases in physical and mental infirmities that interfere with discarding and organizing possessions. Consistent with Wheaton et al. (2008), we also found that the prevalence of hoarding in this community sample was inversely related to household income, even when age, sex, living arrangement, and current employment were controlled statistically. Longitudinal studies are required to determine if hoarding is a response to financial insecurity; or, alternatively, whether financial insecurity is a consequence of hoarding behavior or other characteristics of individuals with this behavior. Third, consistent with prior reports (Frost et al., 2000; Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002; Samuels, Bienvenu, et al., 2007), we found evidence of more impairment in individuals with hoarding behavior. As measured by the GAF, these individuals had poorer psychosocial functioning than individuals who did not hoard. Moreover, we found that the odds of hoarding increased with the number of traits of specific personality disorders, including paranoid, schizotypal, avoidant, and obsessive–compulsive. Consistent with Wheaton et al. (2008), we also found a strong relationship between alcohol dependence and hoarding. We speculate that heavy alcohol use might interfere with discarding and organizing possessions; alternatively, both alcohol dependence and hoarding behavior might be independent outcomes of psychopathology, personality disorders, or adverse life events. Fourth, several self-reported childhood adversities were associated with hoarding in this sample, specifically, lack of security from home break-ins and excessive physical discipline. Interestingly, patients with hoarding sometimes identify break-ins (not necessarily in childhood) as instigating their hoarding behavior (Steketee & Frost, 2007). Hartl et al. (2005) found that excessive physical discipline was strongly correlated with hoarding. We also found that parental psychiatric symptoms (mania, depression, and heavy alcohol use) were associated with hoarding. Parental psychopathology might contribute to childhood adversities, such as material deprivation or excessive physical discipline, that influence the development of hoarding behavior; alternatively, there may be a direct genetic relationship between psychopathology in parents and offspring, such as alcohol dependence, that contribute to hoarding (Nurnberger et al., 2004). The results of the regression analyses in the current study suggest that these four childhood adversities are each independently related to hoarding. We propose that these adversities, and perhaps others not measured in the current study, may be different features of a more general chaotic upbringing, and that some individuals with this history may seek security in collecting and saving a large number of possessions. Alternatively, it has been hypothesized that strong emotional attachment to possessions may be a response to poor attachment to parents during childhood (Steketee, Frost, & Kyrios, 2003). In most previous studies of hoarding, the majority of participants have been women. However, we found a twofold higher prevalence of hoarding in men than women in this community sample. We also found that the magnitude of the relationship between certain correlates and hoarding was different in men and women in this sample. Specifically, parental psychiatric symptoms, and avoidant personality disorder dimensions, were more strongly associated with hoarding in women than men. This may suggest that the development of hoarding may be different in men than women. Wheaton et al. (2008) recently reported that, among women with OCD, those with hoarding had an earlier age at onset of OCD, more severe OCD, and greater prevalence of several Axis I disorders, compared with those without hoarding; whereas the only difference in men with OCD was increased prevalence of social phobia in those who hoarded. In contrast to Wheaton et al. (2008), we found that the magnitude of the association between alcohol dependence and hoarding was similarly strong in men and women in this community sample.

ARTICLE IN PRESS
J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844 843

Strengths and limitations Strengths of the current study include its investigation of hoarding in a community sample; assessment of personality disorder features and specific childhood adversities by psychologists; and evaluation of Axis I disorders, social functioning, and psychosocial stressors by psychiatrists. However, several potential limitations of the study must be acknowledged. First, whereas the 1981 ECA study involved a probabilistic sample of household residents of the eastern Baltimore community, the sample for the current study had undergone multiple sources of attrition by 1997; nevertheless, the gender and ethnic distributions of the participants were similar to those who could not be interviewed (Samuels, Bienvenu et al., 2002; Samuels, Eaton et al., 2002). Moreover, given the relatively low prevalence of hoarding and many of the childhood adversities, a larger sample is required to evaluate the relationships between them with adequate statistical power, as well as to control for multiple correlates simultaneously in regression models. Thus, the findings of this study must be considered preliminary and contingent on replication in larger, representative samples in other communities. Second, we do not know what proportion of individuals with hoarding behavior, as identified in this study, meet criteria for compulsive hoarding as defined by Frost and Hartl (1996), which include acquisition and failure to discard possessions that seem useless or of limited value; clutter that interferes with normal use of living spaces; and significant distress and impairment. In particular, the OCPD hoarding trait used to assess hoarding in this study does not evaluate the extent of clutter, and it focuses on ‘worn-out or worthless objects,’ so that it might not identify some individuals with low insight into their behavior. However, all of the individuals with hoarding either acknowledged, or were judged by the interviewers, to have experienced difficulties because of their hoarding behavior; in addition, in the opinion of several psychiatrists who reviewed transcripts of the cases, the individuals with hoarding behavior in this study were consistent with cases of hoarding that they had seen in clinical and research settings. Third, childhood adversities in this study were evaluated retrospectively. Since the age at onset of hoarding behavior and childhood adversities were not established, it is unknown if specific adversities preceded the onset of hoarding behavior, or vice versa. Moreover, the perception and attribution of certain childhood experiences, as well as the willingness to report them to an interviewer, might be influenced by current psychopathology or personality disorder features (Brewin, Andrews, & Gotlib, 1993). However, the magnitude of the relationships between childhood adversities and hoarding in this study did not appreciably change after controlling for these potential confounders or mediators. Nevertheless, there may be other unmeasured factors that could influence self-report of childhood adversities. Given these difficulties, longitudinal studies are needed that use more valid, reliable methods to evaluate childhood experiences prior to the onset of hoarding behavior, as have been conducted for other mental health outcomes (Fergusson, Boden, & Horwood, 2007; Jaffee et al., 2002). Implications This study provides evidence that hoarding behavior is more prevalent in the community than previously thought. The prevalence is greater in older individuals, and those with limited household income, and these demographic characteristics should be considered in focusing community interventions. Clinicians and social welfare professionals also should be aware that alcohol dependence and paranoid, avoidant, and obsessive–compulsive personality disorder features may complicate the treatment of, and intervention against, hoarding in the community. Moreover, adversities experienced in childhood may independently contribute to the development of hoarding behavior. Further research on the impact of these adversities on behavior may provide insights for developing programs for treating and preventing severe hoarding in the community.

Acknowledgments This research was supported by National Institute of Mental Health Grants R01- MH47447, R01-MH50616, and K23-MH64543. The authors thank the many individuals participating in the study; Ms. Gail Bendit, Ms. Stacie Johnson, Ms. Carolyn Newcomb, and Ms. Valerie Wadja-Johnson for conducting personality assessments; Dr. Leigh Ellison, Dr. David Gotlib, and Dr. William Howard for conducting clinical evaluations; Dr. Armand Loranger and Dr. Alan Romanoski for consultation on the study; and Ms. Margaret Dees, Ms. Sandra Hensley, Ms. April Ravert, and Ms. Krista Vermillion for their efforts in coordinating the study. References
American Psychiatric Association (1987). Diagnostic and statistical manual of mental disorders (3rd ed. revised). Washington, DC: APA. American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: APA. Anderson, S. W., Damasio, H., & Damasio, A. R. (2005). A neural basis for collecting behavior in humans. Brain, 128, 201–212. Anthony, J. C., Folstein, M., Romanoski, A., Von Korff, M. R., Nestadt, G., Chahal, R., et al. (1985). Comparison of the lay Diagnostic Interview Schedule and a standardized psychiatric diagnosis. Archives of General Psychiatry, 42, 667–675.

ARTICLE IN PRESS
844 J.F. Samuels et al. / Behaviour Research and Therapy 46 (2008) 836–844

Brewin, C. R., Andrews, B., & Gotlib, I. H. (1993). Psychopathology and early experience: A reappraisal of retrospective reports. Psychological Bulletin, 113, 82–98. Cromer, K. R., Schmidt, N. B., & Murphy, D. L. (2007). Do traumatic events influence the clinical expression of compulsive hoarding? Behaviour Research and Therapy, 45, 2581–2592. Dykens, E., & Shah, B. (2003). Psychiatric disorders in Prader-Willi syndrome. CNS Drugs, 17, 167–178. Eaton, W. W., Anthony, J. C., Gallo, J., Cai, G., Tien, A., Romanoski, A., et al. (1997). Natural history of Diagnostic Interview Schedule/DSM-IV major depression: The Baltimore Epidemiologic Catchment Area follow-up. Archives of General Psychiatry, 54, 993–999. Fergusson, D. M., Boden, J. M., & Horwood, L. J. (2007). Exposure to single parenthood in childhood and later mental health, educational, economic, and criminal behavior outcomes. Archives of General Psychiatry, 64, 1089–1095. Frost, R., & Gross, R. (1993). The hoarding of possessions. Behaviour Research and Therapy, 31, 367–382. Frost, R., & Hartl, T. (1996). A cognitive–behavioral model of compulsive hoarding. Behaviour Research and Therapy, 34, 341–350. Frost, R. O., Steketee, G., & Williams, L. (2000). Hoarding: A community health problem. Health and Social Care in the Community, 8, 229–234. Frost, R. O., Steketee, G., Williams, L., & Warren, R. (2000). Mood, disability, and personality disorder symptoms in hoarding, obsessive compulsive disorder, and control subjects. Behaviour Research and Therapy, 38, 1071–1082. Goodman, W. K., Price, L. H., Rasmussen, S. A., Mazure, C., Fleischmann, R. L., Hill, C. L., et al. (1989). The Yale–Brown Obsessive Compulsive Scale: I. Development, use, and reliability. Archives of General Psychiatry, 46, 1006–1011. Gothelf, D., Presburger, G., Zohar, A. H., Burg, M., Nahmani, A., Frydman, M., et al. (2004). Obsessive–compulsive disorder in patients with velocardiofacial (22q11 deletion) syndrome. American Journal of Medical Genetics, Part B (Neuropsychiatric Genetics), 126, 99–105. Greenberg, D., Witzum, E., & Levy, A. (1990). Hoarding as a psychiatric symptom. Journal of Clinical Psychiatry, 51, 417–421. Hartl, T. L., Duggany, S. R., Allen, G. J., Steketee, G., & Frost, R. O. (2005). Relationships among compulsive hoarding, trauma, and attention-deficit/ hyperactivity disorder. Behaviour Research and Therapy, 43, 269–276. Jaffee, S. R., Moffitt, T. E., Caspi, A., Fombonne, E., Poulton, R., & Martin, J. (2002). Differences in early childhood risk factors for juvenile-onset and adultonset depression. Archives of General Psychiatry, 59, 215–222. Karno, M., Golding, J. M., Sorenson, S. B., & Burnam, M. A. (1988). The epidemiology of obsessive–compulsive disorder in five US communities. Archives of General Psychiatry, 45, 1094–1099. Loranger, A. W., Sartorius, N., Andreoli, A., Berger, P., Buchheim, P., Channabasavanna, S. M., et al. (1994). The international personality disorder examination. Archives of General Psychiatry, 51, 215–224. Nurnberger, J. I., Jr., Wiegand, R., Bucholz, K., O’Connor, S., Meyer, E. T., Reich, T., et al. (2004). A family study of alcohol dependence: Coaggregation of multiple disorders in relatives of alcohol-dependent probands. Archives of General Psychiatry, 61, 1246–1256. Rasmussen, S. A., & Eisen, J. L. (1992). The epidemiology and differential diagnosis of obsessive compulsive disorder. Journal of Clinical Psychiatry, 53(suppl.), 4–10. Samuels, J., Bienvenu, O. J., III, Riddle, M. A., Cullen, B. A. M., Grados, M. A., Liang, K. Y., et al. (2002). Hoarding in obsessive compulsive disorder: Results from a case-control study. Behaviour Research and Therapy, 40, 517–528. Samuels, J., Eaton, W. W., Bienvenu, O. J., 3rd, Brown, C. H., Costa, P. T., Jr., & Nestadt, G. (2002). Prevalence and correlates of personality disorders in a community sample. British Journal of Psychiatry, 180, 536–542. Samuels, J. F., Bienvenu, O. J., Pinto, A., Fyer, A. J., McCracken, J. T., Rauch, S. L., et al. (2007). Hoarding in obsessive–compulsive disorder: Results from the OCD Collaborative Genetics Study. Behaviour Research and Therapy, 43, 673–686. Samuels, J., Shugart, Y. Y., Grados, M. A., Willour, V. L., Bienvenu, O. J., Greenberg, B. D., et al. (2007). Significant linkage to compulsive hoarding on chromosome 14 in families with obsessive compulsive disorder. American Journal of Psychiatry, 164, 493–499. Saxena, S. (2007). Is compulsive hoarding a genetically and neurobiologically discrete syndrome? Implications for diagnostic classification. American Journal of Psychiatry, 164, 380–384. Saxena, S., Brody, A. L., Maidment, K. M., Smith, E. C., Zohrabi, N., Katz, E., et al. (2004). Cerebral glucose metabolism in obsessive–compulsive hoarding. American Journal of Psychiatry, 161, 1038–1048. Steketee, G., & Frost, R. O. (2003). Compulsive hoarding: Current status of the research. Clinical Psychology Review, 23, 905–927. Steketee, G., & Frost, R. O. (2007). Compulsive hoarding and acquiring: Therapist guide. New York: Oxford University Press. Steketee, G., Frost, R. O., & Kim, H.-J. (2001). Hoarding by elderly people. Health and Social Work, 26, 176–184. Steketee, G., Frost, R. O., & Kyrios, M. (2003). Cognitive aspects of compulsive hoarding. Cognitive Therapy and Research, 27, 463–479. Tolin, D. F., Frost, R. O., Steketee, G., & Fitch, K. E. (2008). Family burden of compulsive hoarding: Results of an internet survey. Behaviour Research and Therapy, 46, 334–344. Wheaton, M., Cromer, K., LaSalle-Ricci, V. H., & Murphy, D. (2008). Characterizing the hoarding phenotype in individuals with OCD: Associations with comorbidity, severity and gender. Journal of Anxiety Disorders, 22, 152–243. Wing, J. K., Babor, T., Brugha, T., Burke, J., Cooper, J. E., Giel, R., et al. (1990). SCAN: Schedules for clinical assessment in neuropsychiatry. Archives of General Psychiatry, 47, 589–593. Wu, K., & Watson, D. (2005). Hoarding and its relation to obsessive–compulsive disorder. Behaviour Research and Therapy, 43, 121–897.