The predictive value of different reasons for saving and acquiring on hoarding disorder symptoms

Abstract

Few studies to date have studied the particular reasons endorsed by individuals with hoarding disorder (HD) for why they save or acquire certain objects. Understanding how reasons for saving and acquiring objects influence hoarding severity and the degree to which the relationship depends on gender or age differences can have implications for the treatment of HD. The current study looked at reasons for saving and acquiring in 84 individuals diagnosed with HD. Consideration of the usefulness of an object as a reason for saving was the most consistently uniquely predictive of all of the reasons examined for saving and acquisition when controlling for gender differences and other endorsed reasons. These results may suggest that targeting specific reasons for saving and acquiring may be an efficient way to reduce hoarding severity, specifically related to ideas of utility and waste.

1. Introduction

Individuals diagnosed with hoarding disorder (HD) are characterized by urges to save objects, difficulty discarding current possessions, and homes with rooms cluttered as to prevent their intended use (American Psychiatric Association, 2013). One aspect of HD that is little understood is the relationship between hoarding severity and the reported reasons those with HD endorse for saving or acquiring objects. Often, hoarding studies have speculated the reasons for saving and acquiring objects, such as that an object contains important information or would otherwise be useful (Steketee, Frost, Wincze, Greene, & Douglass, 2000). Few investigations to date have studied the particular reasons for saving or acquiring objects in this population. Of those that have examined individual rationale for saving/acquiring, most utilized samples with “compulsive hoarding” or “hoarding symptoms,” and not individuals meeting the DSM-5 criteria for HD. Further, many utilized non-clinical, college-aged samples. The following study will explore reasons for saving/acquiring in a formally diagnosed HD sample.

Investigations have shown a link between the number of reasons to save or acquire items in both HD and non-clinical samples. In the early cognitive-behavioural conceptualization of compulsive hoarding symptoms, it was believed that those with hoarding saved items due to sense of responsibility (being prepared and preventing harm to object), sentimental attachments, or for instrumental reasons (such as future need) (Frost & Gross, 1993; Frost & Hartl, 1996; Frost, Hartl, Christian, & Williams, 1995). These initial cited reasons for saving have held up over the years with subsequent investigations and continue to be further refined.

A study on reasons for saving in college students found that a higher endorsement of reasons for saving newspapers and magazines was strongly related to the perceived value of the periodicals and negatively associated with the intentions of the students to discard the magazines and newspapers (Frost, Steketee, Tolin, & Renaud, 2008). Among this nonclinical sample, higher endorsement of reasons for saving and the perceived value of periodicals were both significantly associated with hoarding symptom severity. Frost et al. also found that 14 self-identified individuals with compulsive hoarding reported more reasons for saving objects than did 13 non-clinical individuals, but the specific reasons endorsed by the hoarding group were not identified. The self-identified HD individuals also perceived items as having higher value than did their non-hoarding counterparts (Frost et al., 2008), although no conclusions can reliably be made as to whether individuals with hoarding disorder would be able to generate more reasons to save an object because of the higher value that they place on it.

The Savings Cognition Inventory (SCI; Steketee, Frost, & Kyrios, 2003) has been utilized to investigate beliefs about items when considering whether or not to throw them away. The four categories of reasons for saving examined in the SCI include emotional attachment, memory, control, and responsibility. Individuals with compulsive hoarding (with and without co-morbid OCD) report significantly higher cognitions about saving items (as measured by the SCI) than do non-hoarding controls (with and without OCD), with reasons involving emotional attachment being endorsed most highly by all groups (Gordon, Salkovskis, & Oldfield, 2013; Steketee et al., 2003). Individuals meeting the proposed DSM-5 criteria for HD have also been found to rate items on the SCI higher than self-identified “collectors,” with both groups scoring highest on the emotional subscale (Nordsletten, Fernandez de la Cruz, Billoti, & Mataix-Cols, 2013). Finally, Reid et al. (2011) examined the relationship between reasons for saving and hoarding severity in a sample of community-dwelling, non-hoarding older adults. Higher scores on a revised version of the SCI correlated moderately with the Saving Inventory-Revised (SIR; Frost, Steketee, & Grisham, 2004) total score and all three subscales (Difficulty Discarding, Acquisition, and Clutter). Thus, the frequency and strengths of beliefs about objects is associated with hoarding severity. This suggests that, in general, individuals can recall more instances of having saved an item because of an emotional reason than because of memory issues, feelings of control, or responsibility toward the object. Further, individuals who save items more (i.e., self-identified compulsive hoarding individuals) are also more likely to endorse having reasons to save items.

Grisham et al. (2009) examined reasons for saving and hoarding severity as predicting attachment to objects in self-reported OCD patients with clinically severe hoarding problems. Although both hoarding severity and higher SCI scores were significantly related to the attachment endorsed for a newly acquired object, only the level of attachment at the time of acquirement significantly predicted the level of attachment to the object one week later (Grisham et al., 2009). Although individuals with hoarding problems may initially feel emotionally attached to an object based on their abilities to generate reasons to save it, their attachment to the object at a later time may not be predicted by either the amount of reasons endorsed about why they may wish to save it or their overall hoarding symptoms. Grisham et al. (2009) did not explore which particular items in the SCI predicted attachment of the newly acquired object. Further, the analyses were conducted on individuals with hoarding symptoms as a result of OCD, not HD.

Although the SCI represents an already validated measure assessing reasons for saving, the SCI focuses more on reasons representing non-normal relationships with objects (e.g., “I am responsible for the well-being of this possession;” Steketee et al., 2003) and not on more general reasons for saving or acquiring possessions that are more likely to be endorsed by non-clinical individuals. This represents a major limitation of the SCI because it is not able to assess reasons for saving or acquiring that represent more normal relationship with objects, such as that an object might possess important information or have sentimental value. Further, the SCI assesses reasons for saving objects through specific but indirect items (e.g., “Throwing away this possession is like throwing away a part of me;” Steketee et al., 2003) instead of directly querying participants if they save items for a particular broad category of reasons (e.g., emotional attachment), which might manifest in different specific cognitions in different participants. A further limitation of the SCI is that the SCI is limited to why patients chose to save an object, not why they chose to acquire it in the first place. Although individuals with HD often acquire objects passively (e.g., junk mail or empty food containers), some participants diagnosed with HD may also have problems with excessive acquisition of new possessions (American Psychiatric Association, 2013). There has not yet been a systematic investigation of the possible reasons that individuals with HD might choose to excessively acquire objects. Thus, there is a need to investigate the levels with which individuals with HD endorse different general categories of reasons for saving and acquiring objects.

Nordsletten et al. (2013) queried 29 individuals meeting the proposed DSM-5 criteria for HD and 20 self-identified collectors about specific reasons they might have difficulty discarding possessions (e.g., Useful in future, Sentimental attachment, Monetary value, Avoid waste, Object is unique, Misuse of personal info, Taught to save, Item part of personal identity, Fear bad consequence) and the reasons they acquire new items (e.g., Useful in future, Compelled to acquire it, Fear bad consequence, Compulsive shopper). The most frequently endorsed reason for acquisition or difficulty discarding in both samples was the idea that an object might be useful in the future (Nordsletten et al., 2013). The only items to differentiate collectors from individuals with HD were “Avoid waste” and “Misuse of personal info” as reasons for difficulty discarding as well as “Useful in future” and “Compulsive shopper” as reasons for acquisition. Unfortunately, Nordsletten et al. (2013) did not examine the degree to which endorsement of each item predicted hoarding severity, nor did they control for demographic variables such as gender.

A recent presentation at the Association for Behavioral and Cognitive Therapies (Sinopoli, Frost, Tolin, and Steketee, 2010) examined the influence of endorsement of different reasons for saving or acquiring objects on hoarding severity in individuals with compulsive hoarding, as well as in individuals with OCD or with no clinical diagnoses. Based on prior clinical experiences. Sinopoli, Frost, Tolin, and Steketee (2010) queried participants on the frequency with which they saved items because of reasons of information, emotional attachment, avoiding waste, or aesthetics. Their sample was mostly female (68%) and middle-aged (48.7%). Participants with compulsive hoarding problems more highly endorsed significantly more reasons for saving than the OCD or control groups, and there was no difference in the frequency of reasons endorsed between the OCD and the control group. Individuals with compulsive hoarding reported that avoiding waste was the most frequent reason for saving an object, while the OCD and community control groups reported emotional reasons as their most frequent reason for saving objects (Sinopoli et al., 2010). Within the compulsive hoarding group, when controlling for all other reasons for saving, both avoiding waste and information were significantly predictive of the SI-R total; information, emotional reasons, and avoiding waste were significantly predictive of the SI-R Difficulty Discarding; and only avoiding waste was significantly predictive of both the SI-R Clutter and the SI-R Acquisition. Unfortunately, no information is available about how the sample was recruited or diagnosed and so generalizations cannot be made about whether or not the participants in the study would have met DSM-5 criteria for HD.

With respect to compulsive buying, a phenomenon closely related to HD, individuals report that the acquisition of objects would compensate, reward, or neutralize negative feelings (Krios, Frost, & Steketee, 2004). Individuals suffering from compulsive buying report similar reasons for acquiring items compared to compulsive hoarding samples, including emotional attachment and feelings of security (e.g., Frost & Gross, 1993; Frost & Hartl, 1996; Frost et al., 1995). Further, those with self-reported compulsive buying endorsed greater concerns about maintaining control over their buying and choices. However, beliefs about the power of buying to reduce negative emotions, emotional attachments, concerns about lost opportunities, and beliefs of uniqueness of the purchase were more strongly related to actual buying behaviour than concerns about control.

The current study aimed to further research on reasons for saving and acquiring in individuals with HD in two major ways: (1) examine any differences in level of endorsement and predictive influence of different reasons for saving and acquiring on hoarding severity in individuals diagnosed with HD; and (2) examine the unique predictive values of different reasons for saving and acquiring on hoarding severity while controlling for the possible influence of any gender or age differences in either hoarding severity or reasons for saving and acquiring endorsed by individuals with HD.

Because the diagnosis of HD was only recently introduced by the DSM-5, it is important to scrutinize the generalizability of research conducted with samples of self-reported compulsive hoarders or samples that were evaluated in person but which would not meet the DSM-5 criteria for HD. The participants used in the analyses were interviewed by masters-level graduate research assistants and diagnosed with HD only after a thorough review of the assessment materials by a licensed clinical psychologist, who employed the proposed (and, since May 2013, published) DSM-5 criteria in clinical decisions.

The second aim of the current investigation is to examine the possible effect of gender or age on the predictive abilities of reasons for saving or acquiring on HD symptom severity. Age and gender differences in reasons for saving and acquiring have not been examined in previous studies, and so the potential influence of these two demographic factors on the relationship between endorsed reasons for saving and acquiring and HD symptoms is also currently unknown. Understanding how reasons for saving and acquiring objects influence hoarding severity and how the degree to which that relationship depends on gender or age differences can have implications for the treatment of HD by better allowing clinicians to personalize their interventions to the particular needs of the patient.

2. Methods

2.1. Participants

Eighty-four participants were enrolled between July 2008 and January 2014 for either two individual intervention studies for late-life HD (n = 59) or for a group intervention study for mid-life HD (n = 25). The mean age of the overall sample was 63.76 (SD = 8.06; range: 41–86) and the sample was mostly female (75%).

Recruitment for all studies consisted of flyers distributed throughout the VA San Diego Healthcare System, the University of California San Diego, community and recreation centres, senior centres, and local interest groups (e.g. swap meets). Additional recruitment efforts included postings on craigslist, radio advertisements, referrals from family members, healthcare providers, and professionals working closely with hoarding individuals, and community based presentations.

All study protocols were approved by the Institutional Review Board of the University of California, San Diego and by the VA San Diego Healthcare System. Assessments were conducted at the VA San Diego Healthcare System. No monetary compensation was provided for participation, and all participants provided written informed consent. Only the data from participants’ baseline assessments were utilized in the current investigation. Participants were required to not have cognitive impairment, defined by scores below 23 on the Montreal Cognitive Assessment (MoCA; Nasreddine, Phillips, & Bédirian 2005) and to meet the proposed DSM-5 HD criteria, including the exclusion of individuals for whom the hoarding symptoms were the possible result of other organic/mental disorders, such as traumatic brain injury or dementia.

Participants were required to score over 40 on the SI-R and over 20 on the UCLA Hoarding Severity Scale (UHSS; Saxena, Brody, Maidment, & Baxter, 2007) to be considered for diagnosis. The final HD diagnosis was determined by a consensus diagnosis supervised by a licensed clinical psychologist. Finally, participants were required to have HD as the primary diagnosis. Potential co-morbidities were assessed with the Mini-International Neuropsychiatric Interview (M.I.N.I.; Sheehan, Lecrubier, & Sheehan, 1998). Participants who were identified as having comorbid OCD were required to have HD as the primary diagnosis and for their hoarding symptoms to be unrelated to their OCD. One quarter (25.93%) of the participants in the current analyses were identified as having OCD, with HD as the primary diagnosis.

2.2. Measures

As a part of the demographics questionnaire, participants were asked about the degree to which they saved or acquired objects for four reasons (“Losing important information,” “Sentimental or emotionally significant,” “Wasting a potentially useful object” and “Beautiful or aesthetically pleasing”) (see Appendix 1; adapted from Sinopoli et al., 2010). Four questions asked about possible reasons for saving an object and four questions asked about possible reasons for acquiring an object. Each of the eight items was rated on a Likert-type scale from zero (“not at all”) to eight (“extreme”). All questions about reasons for saving or reasons for acquiring were examined individually in the analyses.

The SI-R is a 23-item self-report scale of hoarding severity with three subscales (Difficulty Discarding, Acquisition, and Clutter). Items are rated on a Likert-type scale (0–4) and the items are summed to create a total score, with higher scores indicating greater severity. The SI-R demonstrated adequate reliability in the present study (α = 0.85).

The Clutter Image Rating (CIR; Frost et al., 2008) is a three-item self-report scale of clutter severity in which participants are asked to select which of nine pictures of three rooms in a house (kitchen, living room, and bedroom) most accurately reflect the amount of clutter within those rooms in their own homes. Scores for the three rooms are averaged to create a mean score, with higher scores corresponding to greater clutter. The CIR demonstrated adequate reliability in the present study (α = 0.82).

2.3. Data analysis

All analyses were performed using Stata version 13.0 (StataCorp, 2013). Possible age and gender differences were examined for all variables. Zero-order correlations were examined among the two hoarding severity measures (the CIR and the SI-R) and all reasons for saving and acquiring.

The strength of a potential underlying factor of saving or acquiring was examined using Cronbach’s alpha for the four reasons for saving, the four reasons acquiring, and the eight items combined.

Two series of paired sample t-tests were run to look at potential differences among the four reasons for saving and among the four reasons for acquiring. To adjust for familywise error, the Bonferroni Correction was used to calculate the differences in reasons for saving and acquiring that could conservatively be considered statistically significant. As there were six t-tests each for reasons for saving and reasons for acquiring, respectively, alpha for each t-test was set to 0.0083 (0.05/6 = 0.0083).

Two sets of four multiple regressions (for a total of eight analyses) were performed to test the unique effects of reasons for saving and reasons for acquiring on hoarding severity when controlling for gender. For these analyses hoarding severity was defined as the SI-R total score and the three SI-R subscales (Difficulty Discarding, Acquisition, and Clutter). Each multiple regression used either SI-R total or one of the SI-R subscales as the outcome variable. The first four multiple regressions examined the unique effect of all the reasons for saving predicting SI-R when controlling for the effect of gender. Each regression had the four reasons of saving and gender predicting either the SI-R total or one of its subscales. The second four multiple regressions examined the same outcomes, but using the four reasons for acquiring to predict the SI-R and its subscales when controlling for the effect of gender.

3. Results

Descriptive statistics for all variables are presented in Table 1. Participants reported similar hoarding severity to other HD samples for both the SI-R (Frost et al., 2004) and the CIR (Frost et al., 2008). On average, participants reported reasons for saving and acquiring as “moderate” to “severe”. The CIR was significantly related to the SI-R total (r = 0.433, p < 0.0001), the SI-R Clutter subscale (r = 0.670, p < 0.0001), but not the SI-R Acquisition subscale (r = 0.072, p = 0.53) or the SI-R Difficulty Discarding subscale (r = 0.03, p = 0.82).

Table 1.

Means, standard deviations, and ranges for all variables.

	M	SD	Min.	Max.
SI-R
Total	61.18	10.44	40	89
Difficulty discarding	19.81	4.05	9	28
Acquisition	15.17	4.83	5	28
Clutter	26.10	5.96	12	36
CIR	4.31	1.84	1.33	9
Reasons for saving
Losing important information	6.19	1.54	0	8
Sentimental or emotionally significant	6.26	1.35	2	8
Wasting a potentially useful object	6.19	1.62	0	8
Beautiful or aesthetically pleasing	5.95	1.75	0	8
Reasons for acquiring
Losing important information	5.43	1.72	0	8
Sentimental or emotionally significant	4.98	1.92	0	8
Wasting a potentially useful object	4.86	1.98	0	8
Beautiful or aesthetically pleasing	4.92	2.11	0	8

3.1. Age differences

Age was not related to hoarding severity as measured by the SI-R Total (r = −0.14, p = 0.22), the SI-R Difficulty Discarding subscale (r = −0.08, p = 0.47), the SI-R Clutter subscale (r = 0.05, p = 0.66), or by the CIR (r = 0.06, p = 0.62). However, age was significantly related to the SI-R Acquisition subscale (r = −0.28, p < 0.01): older participants were less likely to endorse excessive acquisition of objects. Age was unrelated to all reasons for saving and acquiring (all ps > 0.05) except and “Wasting a potentially useful object” as a reason for saving (r = −0.28, p < 0.05) and as a reason for acquiring (r = −0.28, p < 0.05), such that as older adults endorse these reasons to a lesser degree than do younger adults.

3.2. Gender differences

Although women reported significantly greater hoarding severity as measured by the SI-R total (t(82) = 2.54, p < 0.01) and the SI-R subscales Difficulty Discarding (t(82) = 1.76, p < 0.05) and Clutter (t(82) = 1.95, p < 0.05), there were no gender differences for hoarding severity as measured by the SI-R Acquisition subscale (t(82) = 1.50, p > 0.05), or by the CIR (t(82) = 0.45, p > 0.05). Paired t-tests revealed gender-based differences in reasons for saving and acquiring (Table 2). Women endorsed higher agreement with “Beautiful or aesthetically pleasing” with respect to both reasons for saving (t(82) = 2.69, p < 0.01) and reasons for acquiring (t(82) = 2.76, p < 0.01) and “Losing important information” as a reason for acquiring (t(82) = 2.12, p < 0.05). There were no gender differences for reasons of “Sentimental or emotionally significant” or “Wasting a potentially useful object” (all ps > 0.05) with respect to acquiring or saving.

Table 2.

Gender-based differences in age, hoarding severity, and reasons for saving and acquiring in a sample of 63 females and 21 males adults with HD.

Age	Mean (SD)		t
	Women	Men
	64.13(8.29)	62.67(7.43)	.72
SI-R
Total	62.82(10.07)	56.33(10.21)	2.54 **
Difficulty discarding	20.26(3.94)	18.48(4.19)	1.76 *
Acquisition	15.63(5.04)	13.81(3.97)	1.50
Clutter	26.94(5.91)	24.05(5.72)	1.95 *
CIR	4.37(1.82)	4.15(1.94)	0.45
Reasons for saving
Losing important information	6.33(1.46)	5.76(1.73)	1.48
Sentimental or emotionally significant	6.33(1.24)	6.05(1.66)	.84
Wasting a potentially useful object	6.29(1.66)	5.90(1.48)	.93
Beautiful or aesthetically pleasing	6.24(1.64)	5.10(1.81)	2.69 **
Reasons for acquiring
Losing important information	5.66(1.61)	4.76(1.89)	2.12 *
Sentimental or emotionally significant	5.03(2.01)	4.81(1.66)	.46
Wasting a potentially useful object	4.92(2.02)	4.67(1.91)	0.51
Beautiful or aesthetically pleasing	5.27(2.07)	3.86(1.93)	2.76 **

^*p < 0.05.

^**p < 0.01.

3.3. Comparisons of endorsement of reasons for saving and acquiring

Most of the correlations among all reasons for saving and all reasons for acquiring were significant, with several notable exceptions (see Table 3). The reason for acquiring “Losing important information” was only significantly related to “Losing important information” as a reason for saving. Acquiring objects because they are “Sentimental or emotionally significant” was significantly related to all reasons for saving except for saving an object because it is “Beautiful or aesthetically pleasing.” Finally, saving objects because they are “Beautiful or aesthetically pleasing was not significantly related to saving objects because of fear of “Wasting a potentially useful object.”

Table 3.

Inter-correlations for reasons for saving and acquiring in adults with HD.

	1	2	3	4	5	6	7
Reasons for saving
1 Losing important information	1
2 Sentimental or emotionally significant	0.467***	1
3 Wasting a potentially useful object	0.503***	0.313***	1
4 Beautiful or aesthetically pleasing	0.294**	0.494***	0.187	1
Reasons for acquiring
5 Losing important information	0.207***	0.061	0.129	0.196	1
6 Sentimental or emotionally significant	0.234*	0.526***	0.196	0.552***	0.434***	1
7 Wasting a potentially useful object	0.333**	0.301***	0.516***	0.387***	0.291**	0.499***	1
8 Beautiful or aesthetically pleasing	0.220*	0.353***	0.377***	0.664***	0.313**	0.573***	0.590***

^*p < 0.05.

^**p < 0.01.

^***p < 0.001.

All eight items were highly inter-correlated (α = 0.82), as were just the four reasons for saving (α = 0.70) and the four reasons for acquiring (α = 0.77).

The frequency with which each item choice was endorsed for all reasons for saving and acquiring is presented in Table 4. No comparisons of the intensity with which each reason for saving or reason for acquiring was endorsed was significant when using the Bonferroni Correction to control for familywise error (all ps > 0.008). In fact only four comparisons unadjusted p values less than.05: For reasons for saving, “Beautiful or aesthetically pleasing” was endorsed less than “Sentimental or emotionally significant” (t(83)–1.78, p = 0.04). For reasons for acquiring, “Losing important information” was endorsed more than all other reasons, including “Wasting a potentially useful object” (t(83) = −2.24, p = 0.014), “Beautiful or aesthetically pleasing” (t(83) = −1.94, p = 0.03), and “Sentimental or emotionally significant” (t(83) = −1.94, p = 0.03).

Table 4.

Frequency of endorsed response options, means, and standard deviations for reasons for saving and acquiring in adults with HD.

	Not at all		Mild		Moderate		Severe		Extreme	M	SD
	0	1	2	3	4	5	6	7	8
Reasons for saving
Losing important information	1.19	0	2.38	0	8.33	13.10	32.14	20.24	24.72	6.19	1.54
Sentimental or emotionally significant	0	0	1.19	1.19	10.71	9.52	32.14	25.00	20.24	6.26	1.35
Wasting a potentially useful object	1.19	1.19	2.38	1.19	5.95	7.14	39.29	19.05	22.62	6.19	1.62
Beautiful or aesthetically pleasing	1.19	0	4.76	3.57	11.90	3.57	35.71	19.05	20.24	5.95	1.75
Reasons for acquiring
Losing important information	1.20	1.20	7.23	3.61	12.05	9.64	42.17	15.66	7.23	5.43	1.72
Sentimental or emotionally significant	1.19	2.38	11.90	3.57	19.05	16.67	26.19	8.33	10.71	4.98	1.92
Wasting a potentially useful object	1.19	4.76	9.52	5.95	23.81	8.33	27.38	9.52	9.52	4.86	1.98
Beautiful or aesthetically pleasing	2.38	2.38	13.10	4.76	21.43	11.90	16.67	15.48	11.90	4.92	2.11

No reported reasons for saving or acquiring were significantly related to the CIR or the SI-R Clutter subscale (all ps > 0.05). However, all reported reasons for saving and acquiring were significantly related to the SI-R total, the SI-R Difficulty Discarding subscale, and the SI-R Acquisition subscale (all ps < 0.05), with the exception of “Losing important information” as a reason for acquiring, which was not significantly related to any hoarding severity measures (see Table 5).

Table 5.

Correlations between reasons for saving and acquiring and the hoarding severity in adults with HD.

	SI-R total	SI-R acquisition	SI-R difficulty discarding	SI-R clutter	CIR
Reasons for saving
Losing important information	0.269*	0.309**	0.389***	−0.044	−0.116
Sentimental or emotionally significant	0.441***	0.454***	0.546***	0.032	0.009
Wasting a potentially useful object	0.430***	0.395***	0.480***	0.106	−0.020
Beautiful or aesthetically pleasing	0.366***	0.348**	0.456***	0.049	0.081
Reasons for acquiring
Losing important information	0.150	0.208	0.156	−0.019	0.040
Sentimental or emotionally significant	0.315*	0.455***	0.410***	−0.096	0.055
Wasting a potentially useful object	0.447***	0.512***	0.540***	−0.001	0.008
Beautiful or aesthetically pleasing	0.432***	0.543***	0.471***	−0.004	0.110

^*p < 0.05.

^**p < 0.01.

^***p < 0.001.

3.4. Reasons for saving and acquiring predict hoarding severity

The first set of multiple regressions looked at the ability of the four reasons for saving to predict SI-R total and the three SI-R subscales, controlling for gender. Gender and all reasons for saving significantly predicted SI-R total (F(5,77) = 8.38, p < 0.0001, R² = 0.352), SI-R Difficulty Discarding (F(5,77) = 12.54, p < 0.0001, R² = 0.449), and the SI-R Acquisition (F(5,77) = 6.60, p < 0.0001, R² = 0.300), but did not significantly predict participant scores on the SI-R Clutter (F(5,77) = 1.22, p =0.31, R² = 0.073).

When controlling for gender and all other reasons for saving, “Wasting a potentially useful object” and “Sentimental or emotionally significant” as reasons for saving predicted significant additional variance in the SI-R Total, and the SI-R subscales for Acquisition and Difficulty Discarding (see Table 6). “Beautiful or aesthetically pleasing” predicted significant additional variance in the SI-R subscale Difficulty Discarding when controlling for gender and all other reasons for saving.

Table 6.

The unique abilities of gender and reasons for saving and acquiring to predict hoarding severity.

	B(S.E.)
	SI-R total	Difficulty discarding	Acquisition	Clutter
Gender+reasons for saving Gender
Gender	−4.54(2.30)	−0.64(0.82)	−0.83(1.11)	−3.06(1.57)
Losing important information	−.79(0.78)	−.02(0.28)	−0.08(0.37)	−0.69(0.53)
Sentimental or emotionally significant	2.32(0.89) *	0.99(0.32) **	1.01(0.43) *	0.24(0.61)
Wasting a potentially useful object	2.25(0.68) **	0.83(0.25) **	0.83(0.14) *	0.59(0.47)
Beautiful or aesthetically pleasing	0.79(0.65)	0.49(0.23) **	0.36(0.31)	−0.06(0.47)
Gender+reasons for acquiring
Gender	−5.36(2.53)*	−1.35(0.93)	−0.80(1.10)	−3.22(1.66)
Losing important information	−0.52(0.67)	−0.27(0.25)	−0.13(0.29)	−0.17(0.44)
Sentimental or emotionally significant	0.52(0.71)	0.40(0.26)	0.45(0.31)	−0.33(0.47)
Wasting a potentially useful object	1.71(0.66) *	0.82(0.24) **	0.66(0.28) *	0.22(0.43)
Beautiful or aesthetically pleasing	0.72(0.68)	0.24(0.25)	0.62(0.30) *	−0.14(0.45)

^*p < 0.05.

^**p < 0.01.

The second set of multiple regressions looked at the ability of gender and four reasons for acquiring to predict SI-R total and the three SI-R subscales. Gender and all reasons for acquiring significantly predicted SI-R Total (F(5,76) = 6.03, p < 0.0001, R² = 0.2841), SI-R Difficulty Discarding (F(5,76) = 8.61, p < 0.0001, R² = 0.362), and SI-R Acquisition (F(5,76) = 8.62, p < 0.0001, R² = 0.362), but did not significantly predict SI-R Clutter (F(5,76) = 0.98, p = 0.44, R² = 0.061).

Unlike the analyses with reasons for saving, when looking at unique predictors of the SI-R total, gender contributed significant additional variance beyond the endorsed reasons for acquiring (see Table 6). When controlling for gender and all other reasons for saving, “Wasting a potentially useful object” as a reason for acquiring predicted significant additional variance in the SI-R Total, and the SI-R subscales for Acquisition and Difficulty Discarding. Endorsing “Beautiful or aesthetically pleasing” as a reason for acquiring predicted significant additional variance in the SI-R Acquisition subscale when controlling for gender and all other reasons for acquiring.

4. Discussion

The current study explored both the influence of age and gender on hoarding severity and the reasons for saving and acquiring endorsed by individuals formally diagnosed with HD. Analyses were also conducted to investigate the unique abilities of individual reasons for saving and acquiring to explain hoarding severity when controlling for the effect of gender.

4.1. Age differences

Although there was no significant relationship between age and hoarding severity as measured by the CIR, the SI-R Total and the SI-R subscales for Difficulty Discarding and Clutter, age was significantly related to the SI-R Acquisition subscale, suggesting that as individuals with HD age they are less likely to continue to acquire new items at an excessive level, despite having similar cognitions about discarding and similar levels of household clutter. This may be due to several age-related reasons, including physical decline that inhibits one’s ability to acquire or a lack of monetary resources from living on a fixed income. While these reasons are speculative, knowledge of age-related differences in acquisition may help clinicians identify treatment targets that may be different across the life span.

Older participants were less likely to endorse “Wasting a potentially useful object” as a reason for saving current possessions and as a reason for acquiring new items. Although individuals with HD may be consistent across the lifespan with regard to the rates with which they refuse to discard objects, their reasons for doing so may alter with age. Our sample primarily consisted of older adults (mean age = 63.76; SD = 8.06). A larger sample that included more midlife HD patients may provide additional information on age and potential cohort differences in motives for saving and acquiring.

4.2. Gender differences

Women reported higher severity of hoarding symptoms on the SI-R Total and the SI-R Difficulty Discarding and Clutter subscales; however, no gender differences were found on the SI-R Acquisition subscale or on the CIR. Although women with HD may have increased difficulty discarding their current possessions, men and women have equal levels of clutter in their homes and equal levels of excessive acquisition. Several studies have demonstrated that participant ratings in the clinic are strongly related to clinician ratings of the participants’ homes (Frost et al., 2008; Tolin, Frost, & Steketee, 2010), indicating that the CIR may be a more objective measure of clutter severity than the SI-R Clutter subscale, which relies on a verbal description of clutter rather than a visual depiction of cluttered rooms.

Past studies have demonstrated a positive correlation between hoarding severity and endorsement of reasons for saving (Reid et al., 2011). Because women endorsed higher severity of self-reported hoarding symptoms on the SI-R, we expected gender differences in the degree to which participants endorsed all of the queried reasons for saving and acquiring. Interestingly, women only endorsed higher agreement with two of the reasons for saving and acquiring (“Losing important information” and “Beautiful or aesthetically pleasing”). Further, men and women reported equal levels of saving or acquiring an item because it was “Sentimental or emotionally significant” or because they did not want to be “Wasting a potentially useful object.” Although it is important to address emotional and use-driven reasons for saving and acquiring in treatment interventions for HD in both men and women, it may be beneficial to additionally emphasize the aesthetic and informational aspects of objects when working with women with HD.

4.3. Comparisons of endorsement of reasons for saving and acquiring

Although previous characterization studies of reasons for saving using the SCI (Gordon et al., 2013; Nordsletten et al., 2013; Steketee et al., 2003) found that participants reported saving items for emotional reasons significantly more than for any other reasons, the current study found that there were no significant differences in the intensity with which participants endorsed reasons for saving or acquiring.

All reasons for saving and acquiring were endorsed in the moderate to severe range. Further, participants in the current study endorsed all reasons for saving with higher frequency than has been found in previous studies of the same reasons in other HD samples (Sinopoli et al., 2010). Not all reasons were significantly related to one another and no reason was endorsed significantly more than any other. The high reliability coefficients derived from the items (all items as well as just the reasons for saving or the reasons for acquiring) suggest that all items might relate to a single factor for increased reasons for saving and acquiring. Further analyses would be needed to determine if the items could potentially function as an overall measure of likelihood to save current possessions and acquire new items, and if such a measure could discriminate individuals with HD from non-clinical samples.

Participants who more highly endorsed any of the reasons for saving and acquiring items (with the exception of “Losing important information” as reason for saving) were more likely to also have more severe hoarding symptoms, as measured by the SI-R total and subscales for Difficulty Discarding and Acquisition, as is consistent with previous studies (Frost et al., 2008). However, there was no significant relationship between any of the reasons for saving and acquiring and either the CIR or the SI-R clutter subscale. This provides further evidence for a possible divide between severity of hoarding-related cognitions (such as is measured with the SI-R and the reasons for saving and acquiring examined in the current study) and more objective measures of hoarding-related symptoms (such as household clutter as measured by the CIR).

4.4. Reasons for saving and acquiring predict hoarding severity

The current study is the first attempt to explore the different predictive abilities of reasons for saving and acquiring on hoarding severity, when controlling for the tendency of people with HD to endorse high numbers of all reasons for saving an object and gender differences. Although previous studies have focused on reasons for saving as predicting hoarding severity (see Sinopoli et al., 2010), this is the first investigation of unique predictive abilities of reasons for acquiring on hoarding severity as well as the first enquiry to control for gender differences. As expected based on the initial analyses, all examined models were significantly predictive of hoarding severity except for the regression models examining the ability of gender and all reasons for saving or acquiring to predict the SI-R Clutter subscale.

Within the context of the regression models, we were able to examine the unique predictive abilities of each of the reasons for saving and acquiring. Consideration of the usefulness of an object as a reason for saving was the most consistently uniquely predictive of all of the reasons examined for saving and acquisition, which is consistent with Sinopoli et al. (2010). Although increased likelihood to save an item because it is “Sentimental or emotionally significant” was uniquely predictive of hoarding cognitions relating to discarding and acquiring, there was no unique explanatory ability of sentimentality as a reason for acquiring an object, suggesting a disconnect between why individuals with HD acquire objects and why they continue to save them.

Further, endorsement of acquiring items due to their beauty was significantly predictive of self-reported acquisition of items, suggesting that participants who acquire items more because they consider the items to be aesthetically pleasing are more likely to have higher overall rates of acquisition. Participants’ endorsement of saving items due to their beauty was only uniquely predictive of their self-reported difficulty discarding items, suggesting that participants who save items more because of the objects’ perceived beauty are also more likely to have difficulty discarding all objects.

Unlike previous studies (e.g., Sinopoli et al., 2010), we did not find that either the fear of losing important information or emotional reasons were significantly predictive of the SI-R Difficulty Discarding subscale or that the fear of losing important information was predictive of the SI-R Total. This discrepancy may be due to our controlling for gender differences in both hoarding and reasons for saving.

Although gender did not have unique explanatory variance of hoarding cognitions in the context of endorsement of reasons for saving, gender differences were significantly predictive of overall hoarding symptoms (measured by the SI-R Total) when controlling for those same reasons for acquiring. One possible explanation for why the queried reasons for acquiring did not control for gender differences in overall hoarding symptoms is that there are additional reasons that patients with HD acquire objects that are more gender specific than those explored in the current study.

4.5. Clinical implications and limitations

The current study may have implications for empirically based treatment interventions for HD. The results suggest that instead of focusing on reducing total cognitions related to saving or acquiring objects, the more efficient approach may be to focus more specifically on thoughts related to saving items because of potential usefulness and, to a lesser extent, because of their sentimental value. Pre-treatment clinical assessment provides details on each patient’s reasons for saving and not acquiring, and knowledge of the reasons with more predictive power would better inform clinicians about which reasons to focus the cognitive portion of an intervention. However, a major limitation of the current study is that the SCI was not administered to participants and so the reasons for saving assessed using the SCI (i.e., emotional attachment, memory, control, and responsibility) could not be directly compared to the reasons used in the current study in terms of their ability to predict hoarding severity. Future studies should be more inclusive of the reasons for saving and acquiring presented to patients in order to better distinguish the unique predictive abilities of each and thus to better inform clinicians about the disproportionate influence of some reasons for saving or acquiring over others in HD symptom severity.

Another major limitation of the current study is that the reasons for saving and acquiring were not presented as part of a validated measure. A more rigorous investigation of the complex relationship between cognitions about saving or acquiring an object would entail first asking open ended questions in order to establish if there are any reasons that might be more important to individuals with HD than those investigated in the current study. Many possible reasons, such as the intrinsic value of an object or the possibility of gifting the object to someone else, were not included in the reasons provided to participants as options to endorse. The result is that although the current study illustrates the predictive abilities of some reasons for saving and acquiring over others, it is possible that reasons not considered in the analyses, such as those from the SCI or other reasons altogether, would control for the same – or more – variance in hoarding-related cognitions in participants with HD. Further, the current study assumed that the reasons for saving would be the same as the reasons for acquiring. It is equally probable that patients would report very different reasons for saving than for acquiring and this probability was not considered in the current investigation. Future studies would greatly benefit from allowing participants to list their own major motivations for saving or acquiring objects and then determining the major reasons for saving or acquiring endorsed by individuals with HD. The current study is an exploratory investigation that may help to better inform future research of possible specific reasons for saving and acquiring and the different predictive abilities of those specific reasons with respect to each other and gender.

Other limitations of the current study include the limited number of middle-aged participants utilized in the analyses, restricted age range of the sample, and the lack of a nonclinical comparison group. Future studies are needed to compare the particular reasons endorsed in nonclinical samples and how the reasons for saving and acquiring in clinical and nonclinical samples alike influence actual behaviours of saving and acquiring. Finally, the reasons for saving and acquiring endorsed by individuals with HD may be more indicative of the value they place on objects when they reflect on what their reasons for saving or acquiring might have been and are not necessarily the reasons that would be endorsed if participants were asked in the moment of acquisition or attempted discard. A more accurate measure of reasons for saving and acquiring may be to query patients in the moment of saving or acquiring an object as to the reason they are unable to discard or not acquire the object.