|Home | About | Journals | Submit | Contact Us | Français|
Compulsive hoarding, found in many patients with obsessive-compulsive disorder (OCD), has been associated with poor response to serotonin reuptake inhibitor (SRI) medications in some reports. However, no prior study has quantitatively measured response to standardized pharmacotherapy in compulsive hoarders. We sought to determine whether compulsive hoarders would respond as well as non-hoarding OCD patients to the SRI, paroxetine.
Seventy-nine patients with OCD (32 patients with the compulsive hoarding syndrome and 47 patients without prominent hoarding symptoms) were treated openly with paroxetine (mean dose 41.6 ± 12.8 mg/day; mean duration 80.4 ± 23.5 days) according to a standardized protocol, from 3/1993 to 7/2005. All subjects were free of psychotropic medication for at least four weeks prior to study entry. No psychotherapy or psychotropic medications except paroxetine were allowed during the study period. Subjects were assessed before and after treatment with the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS), Hamilton Depression Rating Scale (HDRS), Hamilton Anxiety Scale (Ham-A), Global Assessment Scale (GAS), and Clinical Global Impression/Improvement (CGI) scale.
Both compulsive hoarders and non-hoarding OCD patients improved significantly with treatment (p < 0.001), with nearly identical changes in Y-BOCS, HDRS, Ham-A, and GAS scores. There were no significant differences between groups in the proportions of patients who completed or responded to treatment. Hoarding symptoms improved as much as other OCD symptoms.
Compulsive hoarders responded as well to paroxetine treatment as non-hoarding OCD patients, suggesting that SRI medications are effective for compulsive hoarding. Controlled trials of SRI medications for compulsive hoarding are now warranted.
Although standard diagnostic classifications consider obsessive-compulsive disorder (OCD) to be a single diagnostic entity, factor analytic studies of OCD symptoms have identified four principal symptom factors: (1) aggressive, sexual, and religious obsessions with checking compulsions; (2) symmetry obsessions with ordering, arranging, and repeating compulsions; (3) contamination obsessions with washing and cleaning compulsions; and (4) hoarding, saving, and collecting symptoms (Leckman et al., 1997; Summerfeldt et al., 1999; Cavallini et al., 2002). These symptom factors appear to be relatively stable over time (Mataix-Cols et al., 2002a) and show different patterns of genetic inheritance (Leckman et al., 2003), comorbidity (Samuels et al., 2002), and treatment response (Mataix-Cols et al., 1999, 2002b; Alonso et al., 2001; Denys et al., 2003).
Hoarding is defined as the acquisition of, and inability to discard items even though they appear (to others) to have no value (Frost and Gross, 1993). Compulsive hoarding and saving symptoms, found in 18–42% of OCD patients (Rasmussen and Eisen, 1992; Hanna, 1995; Frost et al., 1996; Samuels et al., 2002), are part of a discrete clinical syndrome that also includes indecisiveness, perfectionism, procrastination, difficulty organizing tasks, and avoidance (Frost et al., 1996). Compulsive hoarding is driven by obsessional fears of losing important items that the patient believes might be needed later (Stein et al., 1999) and excessive emotional attachments to possessions (Frost and Gross, 1993). Living spaces become sufficiently cluttered so as to preclude the activities for which they were designed, causing significant impairment in social and/or occupational functioning (Frost et al., 2000; Saxena et al., 2002). OCD patients who have hoarding and saving as their most prominent and distressing symptom dimension of OCD and show the other associated symptoms listed above are thus considered to have the “compulsive hoarding syndrome” (Saxena et al., 2002; Steketee and Frost, 2003).
Some, but not all, studies investigating the influence of OCD symptom factors on treatment response have found that hoarding and saving symptoms were associated with poor response to pharmacotherapy with serotonin reuptake inhibitors (SRIs) and cognitive-behavioral therapy (CBT). A small study of treatment with paroxetine, placebo, or CBT for 38 OCD patients found that non-responders were significantly more likely to have hoarding/saving symptoms than responders (Black et al., 1998). Hoarding/saving symptoms were present in three of the 17 responders (17.6%) but in 14 of the 21 non-responders (66.7%). However, seven of the 21 non-responders were in the placebo group, and the number of these that had hoarding symptoms was not reported. In a retrospective case series, Winsberg et al. (1999) found that only one of 18 compulsive hoarders treated openly with a variety of SRIs had an adequate response, and nine had no response In an analysis of large-scale, controlled trials of SRI treatment for patients with OCD, higher scores on the hoarding symptom dimension predicted poorer response to SRI treatment, after controlling for baseline severity (Mataix-Cols et al., 1999).
However, three subsequent studies that examined OCD symptom factors and treatment response did not confirm this association. Alonso et al. (2001) found instead that sexual/religious obsessions uniquely predicted poorer long-term outcome after SRI treatment of 60 OCD patients, 37 of whom also received CBT. Erzegovesi et al. (2001) found that poor insight and somatic obsessions predicted poor response to treatment with various SRIs in 159 OCD patients. Shetti et al. (2005) found that sexual obsessions, washing compulsions, and miscellaneous compulsions predicted non-response to SRIs. These three studies each found that hoarding/saving symptoms had no significant effect on response to treatment. Thus, it has remained unclear whether compulsive hoarding is a consistent predictor of poor response to standard anti-obsessional medications.
No prior pharmacotherapeutic study has specifically targeted the compulsive hoarding syndrome a priori, quantified symptom improvement with medication treatment in compulsive hoarders, or prospectively compared response in compulsive hoarders versus non-hoarding OCD patients. Moreover, the total number of compulsive hoarders in the prior studies of predictors of response to pharmacotherapy for OCD discussed above is quite small. CBT studies have suggested that poor outcome in hoarders may be due to premature dropout from treatment (Mataix-Cols et al., 2002b; Abramowitz et al., 2003). Therefore, we sought to determine prospectively whether a larger sample of compulsive hoarders would respond as well as non-hoarding OCD patients to the SRI, paroxetine, whether they would have different dropout rates that influenced outcome, and whether the severity of hoarding symptoms would be related to treatment response.
This study was approved by the UCLA Medical Institutional Review Board. Subjects were 79 consecutive adult patients, enrolled in a brain imaging study (Saxena et al., 2004) from March, 1993 to July, 2005, who met DSM-III-R or DSM-IV criteria for OCD. Compulsive hoarders were recruited through flyers and newspaper advertisements that specifically targeted “packrats, hoarders, and clutterers,” while non-hoarding OCD patients were recruited with advertisements describing other common OCD symptoms. The study was approved by the UCLA Medical Institutional Review Board. All subjects gave informed consent after the procedures and possible side effects were explained by a study physician (S.S. or A.L.B.). Diagnoses were made by clinical interview using first DSM-III-R, and then DSM-IV criteria, and confirmed with the Schedule for Affective Disorders and Schizophrenia – Lifetime (Spitzer and Endicott, 1978). The presence or absence of all types of OCD symptoms was assessed for each subject at study entry with the Yale-Brown Obsessive Compulsive Scale (Y-BOCS – Goodman et al., 1989) and its symptom checklist. To be enrolled, patients had to have a Y-BOCS score ≥16.
For each subject, the severity of hoarding/saving symptoms was rated on a 0–4 scale (0 = none, 1 = mild, 2 = moderate, 3 = severe, 4 = extreme), using Rauch et al's (1998) method of rating severity of OCD symptom factors. Patients were diagnosed prospectively with the compulsive hoarding syndrome only if: (a) compulsive hoarding/saving was their most prominent, distressing, and impairing OCD symptom factor, (b) they met the clinical criteria for compulsive hoarding developed by Frost et al. (1996) requiring clutter that precludes use of living spaces and significant functional impairment due to hoarding, and, (c) they had hoarding severity scores ≥3. Of the 79 patients, 32 met these criteria. Subjects with major medical conditions or concurrent Axis I diagnoses other than major depression, dysthymia, and minor tic disorders were excluded. All subjects were free from psychoactive medications for at least four weeks prior to entering the study, and from fluoxetine for at least five weeks.
All patients were treated openly with paroxetine hydrochloride (Paxil) for 10–12 weeks, according to a standardized protocol. Paroxetine was started at 10 mg/day and increased by 10 mg/day increments every four days to a target of 40 mg/day, as tolerated. Patients remained on the target dose for the first eight weeks of treatment. Thereafter, the dose could be increased to a maximum of 60 mg/ day for weeks 8–12, as tolerated. No psychoactive medications except paroxetine were allowed during the study period. Patients were also not allowed to receive CBT or any other treatments for OCD during the treatment period. Symptom severity was rated before and after treatment with the Y-BOCS, Hamilton Depressive Rating Scale (HDRS – Hamilton, 1960a), Hamilton Anxiety Scale (Ham-A – Hamilton, 1960b), Global Assessment Scale (GAS – Endicott et al., 1976), and Clinical Global Impressions/Improvement Scale (CGI), immediately before and after the treatment period. “Responders” to treatment were defined a priori as those who had ≥35% decrease in Y-BOCS score and a CGI rating of “much improved” or “very much improved.” Patients who had a 25–35% decrease in Y-BOCS score were considered “partial responders” to treatment. These criteria were based on consensus operational definitions of treatment response in OCD (Pallanti et al., 2002).
For a subset of 25 patients (19 compulsive hoarders and 6 non-hoarding OCD patients), the severity of compulsive hoarding symptoms was specifically quantified before and after treatment, using the UCLA Hoarding Severity Scale (UHSS – unpublished), a 10-item, clinician-administered scale that assesses the presence and severity of various components of the compulsive hoarding syndrome, including extent of clutter, urges to save items, excessive acquisition, difficulty discarding, social and occupational impairment, slowing, indecisiveness, and procrastination (see Table 1). Scores reflect the average (mean) occurrence of each symptom over the one week prior to and including the time of the interview. The UHSS was designed as a semi-structured interview that allows additional questions for clarification. In general, scores depend on the patient's report, but the final rating is based on the clinical judgment of the interviewer. Scores can range from 0 to 40. This scale was not available until the last four years of the study, so it could only be used for the last 25 study patients enrolled.
The data were first statistically screened for distributional properties, outliers, and missing values. No variables were rejected by this process. Compulsive hoarders (n = 32) were compared to non-hoarding OCD patients (n = 47). Age, baseline symptom severity scores (Y-BOCS, HDRS, Ham-A, GAS, and hoarding severity score), final paroxetine dose, and duration of treatment were compared between groups with a t-test (two-tailed) for independent samples (SPSS version 6.1.2 for Macintosh). The proportions of females, patients with a current comorbid major depressive episode, patients who completed treatment, and responders to treatment in each group were compared with χ2 tests. Response to paroxetine was compared between the two groups with a last observation carried forward (LOCF) analysis of pre- to post-treatment changes on the symptom rating scales, using repeated-measures ANOVA, with age, gender, and treatment duration as covariates. To determine whether hoarding severity was associated with treatment response across the entire sample of patients (n = 79), partial correlations between pre-treatment hoarding symptom factor score (0–4) and pre- to post-treatment changes on the Y-BOCS, HDRS, Ham-A, and GAS were performed, with age, gender, and treatment duration as covariates.
Hoarders were significantly older than non-hoarding OCD patients (p < 0.001) (see Table 2). There was a significantly higher proportion of women in the compulsive hoarding group than in the non-hoarding group (p < 0.001). Nine of the 32 compulsive hoarders (28%) and 21 of the 47 non-hoarding OCD patients (45%) had current comorbid major depressive episodes, two compulsive hoarders and five non-hoarding OCD patients had comorbid tic disorders, and two of the non-hoarding OCD patients had compulsive skin-picking. The prevalence of current comorbid major depression was not significantly different between the two groups. Mean pre-treatment Y-BOCS, HDRS, Ham-A, and GAS scores were each nearly identical in the two groups.
Sixteen patients were able to tolerate 60 mg/day, seven reached a final dose of 50 mg/day, 38 received 40 mg/day, six could tolerate only 30 mg/day, three were on 20 mg/ day, three had doses reduced below 20 mg/day, and six dropped out of the study before their doses were increased beyond 10 mg/day. The most common side effects limiting dose increases were sedation, fatigue, constipation, akathisia, headaches, and sexual side effects. There was no significant difference between groups in dose of paroxetine, but compulsive hoarders had a longer duration of treatment then the non-hoarding OCD group (t = 3.1, df = 63, p = 0.003). Of the 32 compulsive hoarders, 25 (78%) completed treatment, one patient dropped out due to intolerable side effects, one patient was discharged because of hospitalization for heart disease, and five dropped out for unknown reasons and were lost to follow-up. Forty of the 47 non-hoarding OCD patients (85%) completed treatment; one dropped out due to side effects, and five dropped out for unknown reasons and were lost to follow-up. There was no significant difference in study completion rate between groups (Table 1).
Both groups improved with treatment, with highly significant changes in Y-BOCS, HDRS, Ham-A, and GAS scores (see Table 3). Changes in symptom severity were nearly identical in the two groups, with no between-groups differences on any outcome measure. There was no significant difference between groups in the proportion of patients who met criteria for response to treatment (Table 1). Nine of 32 compulsive hoarders (28%) were classified as responders to treatment, and another seven (22%) were partial responders. In the non-hoarding OCD group, 15 of 47 patients (32%) were responders to paroxetine, and seven (15%) were partial responders. Thus, 50% of compulsive hoarders and 47% of non-hoarding OCD patients had at least a partial response to treatment (>25% decrease in Y-BOCS score). Compulsive hoarders who completed treatment (n = 25) improved slightly more than non-hoarding OCD patients who completed treatment (n = 40) (31% vs. 27% mean decrease on Y-BOCS, between-groups difference not significant). In the subgroup of 25 patients in whom hoarding symptoms were assessed separately from other OCD symptoms, UHSS scores decreased significantly with treatment (F = 24.58, df = 1,16, p < 0.001). UHSS scores decreased from 26.0 ± 3.7 to 19.7 ± 5.8 (24% decline) in the 19 compulsive hoarders rated before and after treatment with the UHSS, and from 4.7 ± 3.5 to 2.5 ± 2.2 (46% decline) in the six non-hoarding OCD patients assessed with the UHSS.
There were no significant associations between hoarding severity and response to paroxetine treatment in any outcome measure. Across all 79 subjects, pre-treatment hoarding symptom factor scores were not significantly correlated with pre- to post-treatment change in Y-BOCS scores (partial r = 0.06, df = 61, p = 0.32), HDRS scores (partial r = 0.02, df = 61, p = 0.45), Ham-A scores (partial r = 0.07, df = 61, p = 0.29), or GAS scores (partial r = 0.09, df = 61, p = 0.24).
To our knowledge, this is the first study to measure and quantify response to standardized pharmacotherapy in patients with the compulsive hoarding syndrome. It also contains the largest sample of compulsive hoarders ever studied in a treatment trial. Compulsive hoarders responded equally as well to paroxetine treatment as non-hoarding OCD patients. The two groups had significant and nearly identical degrees of improvement in OCD symptoms, depressive symptoms, anxiety, and overall functioning. The proportions of patients who met criteria for classification as treatment responders in each group were very similar, and study completers in both groups had very similar decreases on the Y-BOCS. Hoarding severity was not associated with response to paroxetine treatment. Our results are consistent with several previous studies that also did not find hoarding/saving symptoms to predict poor response to SRI medications (Alonso et al., 2001; Erzegovesi et al., 2001; Shetti et al., 2005) and suggest that SRIs may be just as effective for compulsive hoarders as for non-hoarding OCD patients. Further, dropout rates were also similar in the two groups, suggesting that, while hoarders may be more likely than non-hoarding OCD patients to drop out of CBT trials (Mataix-Cols et al., 2002b; Abramowitz et al., 2003), they appear as likely to complete pharmacotherapy trials.
The decline in Y-BOCS scores seen in compulsive hoarders might have been due to improvements in their other, non-hoarding OCD symptoms. Because rating scales for measuring hoarding severity were not available when the study began, we were not able to measure improvement in hoarding/saving symptoms separately from other OCD symptoms in all of our subjects. However, in the subset of 25 patients who were assessed with the UHSS before and after paroxetine treatment, compulsive hoarding symptoms improved significantly, with a mean 25% decrease in UHSS scores. This improvement is very similar to the mean 23% decrease in Y-BOCS scores measured in the overall sample, indicating that compulsive hoarding/saving symptoms responded just as well as other OCD symptoms to paroxetine treatment.
This study had several limitations. The sample size may have been too small to detect significant differences between the two groups in treatment response. However, the numeric differences between groups on nearly every symptom measure were negligible, as was the difference in percentage of responders. The study population was a highly selected sample that excluded most comorbid neuropsychiatric disorders and consisted only of patients willing to participate in a brain imaging study. This may have resulted in a higher degree of motivation for treatment in our subjects. But this would apply to hoarders and non-hoarding OCD patients alike, so it is difficult to determine how it would have eliminated any true difference in treatment responsiveness between the two groups. It is possible that the overall response to treatment may have been better if more patients had been able to tolerate higher doses of paroxetine, since higher SRI doses have been found to be superior to lower doses for OCD (Tollefson et al., 1994; Ackerman et al., 1998; Hollander et al., 2003; Ninan et al., 2006). However, there is no evidence to suggest that higher doses of paroxetine would have preferentially improved symptoms more in one group in this study than the other.
Alternatively, the equivalent response of hoarders and non-hoarding OCD patients in the present study may have been due to other factors. While there was no significant difference between groups in the proportion of patients who had current comorbid major depression, there were slightly more in the non-hoarding OCD group. This non-significantly higher prevalence of major depression may have worsened that group's response, although most studies have not found comorbid depression or HDRS scores to predict worse response to SRI pharmacotherapy, and some have even found that higher HDRS scores predict better response to SRIs (Ackerman et al., 1994, 1998; Denys et al., 2003). Furthermore, mean pre-treatment HDRS scores were virtually identical in the two groups, indicating the same level of depressive symptom severity. All Axis I comorbidities other than major depression, dysthymia, and minor tic disorders were excluded from the study. Since compulsive hoarding and OCD are often associated with comorbid depressive and tic disorders, the inclusion/exclusion criteria used in this study may make its results more generalizable to the broader clinical population than studies that excluded these disorders.
In conclusion, the results of this study suggest that SRIs may be as effective for compulsive hoarders as they are for non-hoarding OCD patients. Double-blind, placebo-controlled trials of SRI medications for the compulsive hoarding syndrome are now warranted.
This work was supported by an NIMH grant (R01 MH069433) and an NIMH Career Development Award (K23 MH01694) to Dr. Saxena; a grant from the Obsessive-Compulsive Foundation to Dr. Saxena; a NIMH grant (R01 MH53565A) to Drs. Baxter; grants from the National Alliance for Research in Schizophrenia and Depression to Drs. Brody and Baxter; a Veterans Affairs Type I Merit Review Award to Dr. Brody; a Department of Energy grant (DE FCE3-87ER 60615) to Dr. Baxter; and donations from Mr. and Mrs. Brian Harvey.
This work was presented in part at the Anxiety Disorders Association of America Annual Meeting in Seattle, Washington on March 19, 2005; and at the American College of Neuropsychopharmacology Annual Meeting in Kapalua, HI on December 13, 2005.