Hoarding Symptoms in Late Life Depression are Associated with Greater Executive Dysfunction and Disability and Poorer Response to Depression Treatment

Abstract

Objectives:

Late Life Depression (LLD) and Hoarding Disorder (HD) are common in older adults and characterized by executive dysfunction and disability. We aimed to determine the frequency of co-occurring HD in LLD and examine hoarding severity as an additional contributor to executive dysfunction, disability, and response to psychotherapy for LLD.

Design:

Cross-sectional

Setting:

Outpatient psychiatry program

Participants:

83 community-dwelling adults ages 65–90 with LLD

Intervention:

Problem-Solving Therapy

Measurements:

Measures of executive function, disability, depression, and hoarding severity were completed at post-treatment. Pearson’s Chi-squared tests evaluated group differences in rates of cognitive impairment, disability, and depression treatment response between participants with HD (LLD+HD) and LLD only. Separate linear regressions assessed associations between hoarding severity and executive function, disability, and psychotherapy response. Covariates included age, education, gender, and depression severity.

Results:

30.1% (25/83) of LLD participants met HD criteria. Relative to LLD, LLD+HD participants demonstrated greater impairment rates on measures of executive function (Letter-Number-Sequencing, X²(1)=4.0,p=.045; Stroop-Interference, X²(1)=4.8, p=.028). Greater hoarding severity was associated with poorer executive functioning performance (Letter-Number-Sequencing (t(70)=−2.1,β=−0.05,p=.044), Digit-Span (t(71)=−2.4,β=−0.07,p=.019), Letter-Fluency (t(71)=−2.8,β=−0.24,p=.006)). Rates of disability were significantly higher for LLD+HD (88.0%) than LLD (62.3%), (X²(1)=5.41,p=.020) and higher hoarding severity was related to greater disability (t(72)=2.97,β=0.13,p=.004). Depression treatment response rates were significantly lower for LLD+HD (24.0%) compared to LLD (48.3%), X²(1)=4.26,p=.039, and HD status predicted psychotherapy response, t(67)=−2.15,β=−15.6,p=.035.

Conclusions:

We found 30.1% co-occurrence of HD in LLD, which was accompanied by greater executive dysfunction, disability, and poorer response to depression treatment. Results underscore the need for increased screening of hoarding behaviors in LLD and tailored interventions for this LLD+HD group.

1. Objective

Late Life Depression (LLD), or Major Depressive Disorder (MDD) in older adults, is commonly associated with executive dysfunction (1) and is known to contribute substantially to global disability (2) with estimated prevalence rates of 15–29% among community-dwelling older adults (3, 4). Psychotherapy interventions are effective (5) and commonly used to treat LLD with response rates ranging from 41–57% (6, 7). Hoarding Disorder (HD) is estimated to affect 4–6% of older adults, with increasing diagnostic prevalence with age (8), and is also significantly associated with executive dysfunction (9) and disability (10). Despite overlapping neurobiological features of dysfunction in HD and LLD, the concurrence of HD in primary presenting LLD has been minimally investigated. Additionally, given that executive dysfunction and disability are prominent clinical characteristics of both HD and LLD (1, 11), evaluating the potential for compounded impact on executive dysfunction, disability, and response to depression treatment are critical to improve intervention outcomes and aid in early identification of individuals at risk for cognitive and functional decline.

HD is characterized by persistent difficulty discarding objects which often results in excessive acquisition of items, leading to significant clutter over time (12). Prior to the Diagnostic and Statistical Manual of Mental Disorders (DSM)-5^th edition (12), HD was not recognized as a unique diagnostic entity, somewhat limiting research efforts historically. Previously considered a subtype of Obsessive-Compulsive Disorder (OCD) (13) or a symptom of Obsessive-Compulsive Personality Disorder (13), HD concurrence with OCD is now recognized as relatively infrequent at 6–18% (14–16), particularly in comparison to HD comorbidity with depression (14, 15, 17). Among individuals with HD, lifetime history of MDD is estimated to be approximately 50% (15, 16). However, only one study to-date has specifically examined the prevalence of HD in individuals with primary presenting MDD, which documented comorbid HD in nearly 8% of middle-aged adults (18). Yet, despite evidence for increased prevalence of HD in older age (8), the concurrence of HD in older adults with depression (i.e., LLD) remains under-explored.

Deficits in executive functioning are a prominent feature of cognitive dysfunction in LLD (1) and have been linked to poorer treatment response (19) and greater disability (20). While more limited information is available for HD, cognitive dysfunction is reported in up to 90% of individuals with HD (21) and has been similarly described in domains of executive functioning, particularly on visually-mediated tasks (9, 11, 22–24). The sole study that has investigated significant hoarding behaviors in the context of LLD found greater executive dysfunction for categorization and problem-solving in the LLD group with severe compulsive hoarding behaviors relative to LLD alone (22), though limited by a small sample size. Further, in older adults with HD, Ayers and colleagues found that hoarding symptom severity predicted disability, beyond the contribution of depression and anxiety (25).

The objectives of the present study were to determine the frequency of co-occurring clinically significant hoarding symptoms and to examine the associations of hoarding symptom severity with executive functioning, disability, and response to psychotherapy in a sample of older adults presenting for psychotherapy treatment for LLD. We hypothesized that concurrence of clinically significant hoarding in LLD would be higher than rates typically reported in OCD (≤18%) or among the general population of older adults (6%). Given the reported deficits in executive functioning and disability in HD, we also expected clinically significant hoarding symptoms in LLD to be associated with greater executive dysfunction and disability, and poorer response to psychotherapy treatment for depression relative to LLD alone.

2. Methods

2.1. Participants

Primary English-speaking adults aged 65 and older were recruited from the local community for participation in a psychotherapy treatment study for LLD. Inclusion criteria required all participants to meet diagnostic criteria for unipolar MDD without psychotic features according to the Structured Clinical Interview for the DSM-4^th edition (SCID-IV), current depressive episode duration ≥6 weeks, and moderate to severe depression (≥19 on the 24-item GRID-Hamilton Depression Rating Scale (HDRS)) (26, 27). Participants were excluded on the basis of antidepressant use or psychotherapy within the past 6 weeks; electroconvulsive therapy within the past 6 months; use of cognitive enhancing medication; comorbid primary psychiatric diagnoses characterized in the DSM-IV aside from Generalized Anxiety Disorder or Specific Phobia; neurological or major medical conditions; history of substance use disorder within the past 6 months; traumatic brain injury or loss of consciousness ≥30 minutes; and history of diagnosis or current evidence of dementia (e.g., Mini Mental Status Exam score <26). Eligibility for study participation was confirmed via multidisciplinary consensus for each participant.

2.2. Procedure

The study was approved by the University of California, San Francisco Institutional Review Board and conducted in accordance with the Declaration of Helsinki. All eligible participants provided written informed consent. Participants completed measures of executive function, disability, and depression and hoarding symptom severity at pre- and post-treatment visits for a 12-session course of Problem-Solving Therapy (PST) (28) for LLD treatment (7, 29) with a licensed clinical psychologist trained in the PST protocol. Demographic information (e.g., participant age, gender, years of education) and depression history characteristics (e.g., onset age, current episode length) were collected. The present study evaluating the impact of hoarding severity was conducted as a secondary analysis of a Pre-registered Clinical Trial (https://clinicaltrials.gov/study/NCT02440815) designed to examine change in depression, cognition (30), and neurophysiological markers following PST for LLD. Given that not all participants completed the hoarding severity measure at baseline, our primary outcomes (executive function, disability, and depression treatment response) were evaluated using post-treatment data. This allowed for maximizing use of outcome data collected concurrent with hoarding symptom severity as well as assessment of treatment response.

2.3. Measures

The Savings Inventory-Revised (SI-R) (31, 32), a self-report measure of hoarding symptom severity was administered at both pre-treatment baseline and post-treatment visits, or at post-treatment only. Participants who completed the SI-R and obtained a total score of ≥41 at either baseline or post-treatment were classified as having clinically significant hoarding symptoms (18, 33) concurrent with LLD, reflecting LLD with probable HD, herein referred to as LLD+HD. Recent work suggests that a lower threshold for clinically significant hoarding symptoms (SI-R≥33) may be more appropriate for older adults (34); however, given that a comprehensive clinical assessment for Hoarding Disorder was not conducted as a part of this study, we opted to utilize the more conservative cutoff score (SI-R≥41) to classify participants with clinically significant hoarding symptoms and reduce potential misclassification due to lower specificity in this group. All cognitive and functional data at the post-treatment visit were used in analyses. Participants who completed SI-R at both baseline and post-treatment were classified as LLD+HD if their scores at either timepoint were above the SI-R threshold for clinically significant hoarding. For regression analyses, post-treatment SI-R data were used regardless of diagnostic group classification.

Six measures of executive functioning were administered. The Wechsler Adult Intelligence Scale-III (WAIS-III) Digit Span and Letter-Number Sequencing were used to assess auditory verbal working memory with total correct as the primary outcome variables (35). The Controlled Oral Word Association Test (COWAT), a verbal fluency task for letters ‘F-A-S’ examined total words generated as the primary outcome (36). The Trail Making Test-Part B (TMT-B) evaluated rapid set-shifting with the primary outcome as total completion time in seconds (37). WAIS-III Digit Symbol-Coding was also employed to assess visual information processing speed with an executive function component and used total items correct as the primary outcome variable (35). The Stroop Color-Word Interference Test was administered to evaluate inhibition of a prepotent response with total correct on the inhibition trial as the primary outcome (38). Cognitive data were standardized based on normative information provided in the respective manuals for each task and impairment cutoffs of 1 standard deviation (SD) below the mean were applied.

The World Health Organization Disability Assessment Schedule (WHODAS)-II 12-item version assessed self-reported disability for performing activities of daily living, with higher total scores indicating greater disability (39). Employing a conservative cutoff, participant WHODAS-II total scores <5 were deemed as having little to no impairment in daily functioning while total scores ≥5 were classified as indicating functional impairment (39).

The 24-item GRID-Hamilton Depression Rating Scale (HDRS) (26, 27) assessed depression severity. Depression treatment response was classified as ≥50% improvement on the GRID-HDRS score from baseline to post-treatment, irrespective of HD classification.

All measures were administered by trained research personnel. To leverage the most data available, scores on the SI-R, HDRS, WHODAS-II, and executive functioning measures at post-treatment were used in the primary analyses.

2.4. Statistical Analyses

The concurrence of clinically significant hoarding symptoms (i.e., HD) was assessed at baseline and post-treatment. Pearson’s Chi-squared tests were used to evaluate post-treatment group differences between LLD and LLD+HD regarding rates of executive function impairment, disability, and psychotherapy response. Separate linear regression analyses were computed to examine dimensional associations between hoarding symptom severity (SI-R) and measures of executive functioning and disability (WHODAS-II), and included as covariates participant age, years of education, gender, and concurrent depression severity (HDRS). Exploratory linear regression analyses were performed to further examine the impact of hoarding on disability and treatment response after accounting for group differences in impaired executive functioning to investigate the contribution of executive dysfunction on these outcomes. Raw scores on the WHODAS-II and HDRS percent change scores at post-treatment were used as the criterion variables with age, education, gender, baseline depression severity, cognitive impairment classification, and clinically significant hoarding symptoms as predictors. As raw scores are the inputs for the regressions, the beta (β) regression coefficients are unstandardized.

3. Results

3.1. Prevalence of Significant Hoarding Symptoms in Late Life Depression

Of the 105 enrolled study participants who completed the protocol, 83 participants completed the SI-R at post-treatment and 36 of those individuals completed the SI-R at both baseline and post-treatment. At baseline, 27.8% (10/36) of participants met criteria for significant hoarding symptoms (SI-R≥41). At post-treatment, 24.1% (20/83) met criteria for significant hoarding symptoms and 30.1% (25/83) met criteria for significant hoarding at either the baseline or post-treatment visit. There was no difference between participants who completed SI-R at both baseline and post-treatment (n=36) and those who completed SI-R at post-treatment only (n-47) in age, education, gender, ethnicity, and race distribution.

The LLD groups with (n=25) and without (n=58) significant hoarding symptoms did not differ on age, years of education, gender, ethnicity, baseline depression severity, current MDD episode length or MDD onset age; self-identified race was higher in proportion as “White” and lower in proportion as “Black/African American” among the LLD only group relative to the LLD+HD group (Table 1). Of the 83 participants who completed the SI-R at post-treatment, 77 participants had complete cognitive data, 78 had complete WHODAS-II data, and 83 had complete depression severity ratings (HDRS).

Table 1.

Demographic and depression history characteristics by group

	Mean (SD)		X2 or t(df)	p
Variable	LLD+HD (n=25)	LLD Only (n=58)
Gender, n (%) women	14 (56%)	29 (50%)	X2(1)=0.25	.616
Age, years	72.52 (4.65)	70.84 (5.60)	t(81)=1.31	.193
Education, years	15.68 (2.73)	16.41 (2.16)	t(81)=−1.31	.195
Ethnicity, n (%) Hispanic or Latina/o/x/e	3 (12%)	4 (6.9%)	X2(1)=0.59	.443
Race, n (%)
Asian/Asian American	4 (16%)	6 (10.3%)	X2(1)=0.53	.468
Black/African American	3 (12%)	1 (1.7%)	X2(1)=4.02	.045
White	16 (64%)	49 (84.5%)	X2(1)=4.32	.038
More than one race	2 (8%)	2 (3.4%)	X2(1)=0.79	.374
Baseline MMSE	28.88 (1.24)	29.19 (1.15)	t(81)=−1.10	.273
Baseline HDRS	23.7 (2.72)	24.4 (4.12)	t(81)=0.78	.439
MDD Onset Type†, n (%) early onset‡	20 (80%)	47 (85.45%)	X2(1)=0.38	.540
MDD Onset Age†, years	31.84 (24.2)	28.27 (20.3)	t(78)=0.68	.496
Current Length MDE†, months	176.60 (253.0)	165.16 (257.0)	t(78)=0.19	.854
Baseline SI-R, total n=36	49.7 (12.1), n=12	25.2 (10.4), n=24	t(34)=6.28	<.0001

Note. LLD+HD group includes participants who met criteria for significant hoarding symptoms (SI-R≥41) at any time point.

MMSE=Mini Mental Status Exam; HDRS=GRID-Hamilton Depression Rating Scale; MDD=Major Depressive Disorder; MDE=Major Depressive Episode; SI-R=Savings Inventory-Revised.

^†3 participants had missing data for MDD history characteristics (in the LLD only group).

^‡MDD Onset Type early onset is defined as MDD onset age ≤65 years.

We also examined variation in hoarding severity ratings among the subset of participants for whom this data was available (n=36). Of the 36 individuals who completed SI-R at both baseline and post-treatment, 5 participants met criteria for clinically significant hoarding (i.e., HD) at both baseline (M=57.4, SD=9.3) and post-treatment (M=58.2, SD=9.7); 5 participants met HD criteria at baseline only (M=49.0, SD=7.3) although SI-R ratings remained somewhat elevated at post-treatment (M=35.0, SD=4.0); and 2 participants met HD criteria at post-treatment only (M=49.5, SD=10.6) with elevated ratings at baseline (M=32.0, SD=11.3). These 12 participants who met criteria for HD at either time point evidenced a mean change of −2.58 (SD=13.4) in SI-R score with no significant difference in ratings from baseline to post-treatment (t(11)=0.67, p=.517). Additionally, among all 36 participants, a mean change of −2.44 (SD=9.47) points in SI-R total score was observed over the intervention with no significant difference in pre- to post-treatment ratings (t(35)=1.55, p=.131), together suggesting limited change in hoarding symptom severity during the 12 weeks of psychotherapy for LLD. Moreover, hoarding symptom severity did not significantly change based on eventual depression treatment response status, with a mean change in SI-R total score of −3.15 (SD=8.98) for responders and −2.04 (SD=9.92) for non-responders, t(34)=0.33, p=.741.

3.2. Hoarding Severity in Late Life Depression and Executive Functioning

Pearson’s Chi-squared tests evaluated group differences in impairment rates in executive functioning following treatment (Figure 1A). Participants who met criteria for significant hoarding symptoms at either time point (LLD+HD, n=23) demonstrated greater impairment rates at post-treatment in Letter-Number Sequencing (LLD+HD=13.0% (95% CI=2.8-33.6), LLD=1.9% (95% CI=0.05-10.1), X²(1)=4.0, p=.045) and Stroop Color-Word Interference (LLD+HD=26.1% (95% CI=10.2-48.4), LLD=7.5% (95% CI=2.1-18.2), X²(1)=4.8, p=.028) relative to the LLD only group (n=53). Impairment rates by group (LLD+HD, n=23; LLD only, n=53) were not significantly different for Digit Span (p=.404), Digit Symbol-Coding (p=.235), TMT-B (p=.539), or COWAT (p=.109). Of note, the LLD+HD and LLD only groups did not differ on impairment rates in executive functioning measures at baseline (all p>.05). Regression analyses at post-treatment covarying for participant age, years of education, gender, and concurrent depression severity revealed that greater hoarding symptom severity (SI-R) was significantly associated with poorer executive functioning performance on Digit Span (t(71)=−2.4, β=−0.07, p=.019), Letter-Number Sequencing (t(70)=−2.1, β=−0.05, p=.044), and COWAT (t(71)=−2.8, β=−0.24, p=.006).

Figure 1.

Pearson’s Chi-squared tests evaluated rates of impairment at post-treatment in executive function, disability, and response to psychotherapy treatment for depression by group. (A). Impairment rates in measures of executive function show significantly greater impairment in the LLD+HD group on Letter-Number Sequencing (X²(1)=4.0, p=.045) and Stroop Interference (X²(1)=4.8, p=.028) relative to LLD only. (B). Rates of disability as measured by the WHODAS-II-Short indicate significantly higher rates of functional impairment among the LLD+HD group compared to LLD only (X²(1)=5.41, p=.020). (C). Depression treatment response rates were significantly lower for LLD+HD relative to LLD only (X²(1)=4.26, p=.039).

* p<.05.

3.3. Hoarding Severity in Late Life Depression and Disability

A Pearson’s Chi-squared test indicated that rates of disability following treatment were significantly higher for individuals with clinically significant hoarding (LLD+HD, n=25) at 88.0% (95% CI=68.8-97.5) than LLD only (n=53) at 62.3% (95% CI=47.9-75.2), (X²(1)=5.41, p=.020; Figure 1B). Of note, LLD+HD (96%) also demonstrated higher rates of disability at baseline relative to the LLD (70.7%) group (X²(1)=6.59, p=.010). Results of regression analyses assessing the association between hoarding symptom severity and disability at post-treatment revealed that SI-R significantly predicted degree of disability as measured by the WHODAS-II (t(72)=2.97, β=0.13, p=.004), such that greater hoarding symptom severity was related to higher disability, after accounting for depression severity and demographic covariates. Furthermore, at post-treatment, greater hoarding symptom severity remained associated with higher disability even after accounting for cognitive impairment classification (impaired vs. unimpaired) on the executive functioning tasks for which a significant difference in impairment rates was evident between the LLD and LLD+HD groups, (t(66)=3.05, β=0.14, p=.003).

3.4. Hoarding Severity in Late Life Depression and associations with Treatment Response

Among the full sample of 83 participants who completed the SI-R, 40.96% (34/83) were classified as responders to psychotherapy treatment for depression. A Pearson’s Chi-squared test indicated that depression treatment response rate was significantly lower at 24% (6/25; 95% CI=9.4-45.1) for the LLD+HD group compared to 48.3% (28/58; 95% CI=35.0-61.8) for the LLD only group, X²(1)=4.26, p=.039 (Figure 1C). Linear regression indicated that clinically significant hoarding symptoms were associated with poorer treatment response (i.e., lower percent change in HDRS scores), t(67)=−2.15, β=−15.6, p=.035, even after accounting for group differences in impaired executive functioning and demographic characteristics. Age, education, gender, baseline depression severity, and impaired executive functioning were not significant contributors to psychotherapy response.

4. Conclusions

To our knowledge the present study is one of the first to assess the frequency of co-occurring clinically significant hoarding symptoms in primary presenting LLD. We found a 24–30% concurrence of clinically significant hoarding symptoms in older adults with LLD. Additionally, results indicated that clinically significant hoarding symptoms in LLD were associated with greater executive dysfunction, increased disability, and poorer response to depression treatment compared to LLD alone. These results have important implications for cognitive, functional, and treatment outcomes in LLD with concurrent HD and underscore the need for increased screening of hoarding symptoms in LLD and tailored interventions for individuals with both conditions.

We found that 24–30% of individuals with primary presenting LLD also exhibited clinically significant hoarding symptoms. This concurrence rate for HD with LLD is considerably higher than reported rates for HD concurrence with OCD (6–18%) (14–16) or among the general population of older adults (6%) (8). While routine assessment of co-occurring disorders (e.g., OCD, depression, anxiety disorders) in the context of primary presenting HD is common in clinical settings, older adults presenting with primary MDD in the clinic may be less likely to be screened for hoarding symptoms. Thus, hoarding symptoms are likely underappreciated and consequently undertreated in older adults presenting clinically with primary MDD. Given increases in hoarding behaviors with advancing age (8) and associations of LLD with neurodegenerative disease (40), the potential for hoarding behaviors in LLD to represent a neuropsychiatric feature of a neurodegenerative process (17) warrants further investigation. While there were no differences in distribution of early and late onset MDD between groups with and without clinically significant hoarding symptoms in the present sample, we did not investigate onset age of hoarding symptoms. Additionally, although we report on frequency of co-occurring clinically significant hoarding symptoms in LLD, an important future direction would be to assess the incidence of hoarding symptoms among specific age groups. Moreover, future work examining the course of hoarding symptomatology in LLD, including onset age of hoarding behaviors, is critical and may provide additional information about mechanisms associated with hoarding symptoms in older adults and aberrant executive functioning and disability.

The present results indicated that clinically significant hoarding symptoms in LLD were associated with greater executive dysfunction compared to LLD alone. Higher impairment rates on select measures of executive functioning including auditory working memory and inhibition in those with LLD+HD were observed. In addition, across the sample, greater hoarding symptom severity was associated with poorer auditory working memory and verbal fluency performance. These results align with prior work demonstrating executive dysfunction in LLD (1) and HD (9) separately and provide evidence for a compounded effect on aspects of executive functioning in co-occurring LLD+HD. Although only two measures of executive function assessed indicated group differences in impairment rates using categorical approaches (e.g., impaired vs. unimpaired), our dimensional analyses demonstrated that hoarding severity was associated with executive dysfunction on three measures, including Digit Span, Letter-Number Sequencing, and COWAT. Notably, our impairment cutoffs were based on the normative data from the corresponding test manuals, and therefore derived from different normative samples, posing a limitation as we are unable to assess the similarity of these normative groups with one another.

Importantly, while cognitive dysfunction is highly prevalent in HD alone, current literature on executive functioning in HD is less consistent than in LLD. Although the field is in a nascent stage, the present results suggest that HD in the context of LLD may show differential associations with executive processes, warranting further exploration. Additionally, as we did not include measures independent of processing efficiency, we are unable to comment on other visually-mediated executive functioning tasks such as categorization and problem-solving that have been shown to be impaired in HD alone (23, 24). Further, we found no baseline differences in executive functioning between the LLD and LLD+HD groups. Taken together with our result at post-treatment demonstrating poorer EF in the LLD+HD group, this lack of baseline difference may reflect potential improvement following PST for the LLD group only; however, given that we did not examine longitudinal change in EF, future work will be critical to investigate this possibility. Further, this result is likely compounded by the effect of a smaller sample at pre-treatment, reducing our power for these analyses. Alternatively, it has also been postulated that executive dysfunction contributes to the emergence, maintenance, and potential severity of hoarding behaviors (41), further underscoring the need for longitudinal investigation of the course of hoarding symptomatology and executive dysfunction in LLD.

Participants with LLD+HD also demonstrated higher levels of disability compared to those with LLD only. Such findings indicate that hoarding symptoms in LLD may have an additive effect on disability. While recent work across the age range found higher functional impairment rates in MDD alone compared to HD alone, individuals with comorbid MDD and HD demonstrated the highest functional impairment rate (10), akin to our results in LLD+HD. Similarly aligned is prior work in older adults with HD in which hoarding symptom severity predicted disability, beyond the contribution of depression and anxiety (25). Greater levels of disability may lead to poorer prognosis in this LLD+HD subgroup and place these individuals at risk for more rapid decline. Furthermore, our results indicating that higher hoarding severity remained associated with greater disability beyond the contribution of impaired executive functioning suggest that concurrent HD is driving this relationship while executive dysfunction has limited influence.

Finally, we found that the LLD+HD group evidenced poorer depression treatment response relative to LLD alone. Moreover, results indicated that HD was associated with poorer treatment response beyond the contribution of residual executive functioning impairment in this sample. These findings are particularly important for clinical care, as hoarding symptoms are likely under-evaluated, and patients demonstrating hoarding symptoms in the context of LLD may be less likely to respond to conventional psychotherapy treatments for LLD. Although PST has been linked to positive treatment outcomes in LLD for individuals with executive dysfunction (7), the burden of hoarding may reduce the potential benefits of PST. Given a 30% co-occurrence of HD in LLD, hoarding behaviors may be an important factor to identify individuals who may be more resistant to depression treatment and potentially at greater risk of cognitive and functional decline.

Our study does have limitations and we interpret our results cautiously. Specifically, given that this study was not designed to examine hoarding symptoms as a predictor of response to PST, hoarding symptom severity measures were not obtained for all participants at both timepoints. While our results demonstrate that hoarding symptom severity remained relatively stable throughout the 12-week psychotherapy intervention for the subset of individuals who completed SI-R at both baseline and post-treatment (n=36), we primarily explored associations with executive function and disability at post-treatment of an intervention designed for depression to maximize our statistical power. Our analyses of pre-treatment executive function and disability have smaller sample sizes than at post-treatment and we therefore interpret these results cautiously. Additionally, we did not explicitly examine longitudinal change in executive function or disability with treatment. Further, this was not a randomized controlled trial with a treatment comparison group and therefore we cannot determine whether any change in depression was the direct result of PST. Given that we did not correct for multiple comparisons in this secondary data analysis, we recognize the increased chance of Type I error as a limitation and exercise caution in interpretation. Notably, race distribution was less diverse in the LLD only group relative to the LLD+HD group, which may represent a confound in our results as race was not explicitly accounted for in the analyses. Additionally, the lack of ethno-racial diversity across the entire sample limits our results in that they do not adequately represent, and thus may not generalize, across ethno-racial groups. Moreover, our study examined clinically significant hoarding symptoms in a treatment-seeking sample of individuals with LLD and evaluated only those participants who completed the study protocol, which may not be fully representative of all older adults with MDD. As such, future work in a larger, diverse, and representative sample of community dwelling older adults with MDD is warranted. Nonetheless, a notable strength of this study is that we are able to assess differential treatment response to psychotherapy for LLD with and without HD and suggests that older adults with concurrent HD and LLD may be particularly resistant to psychotherapy treatment. Given the rate of concurrent symptomatology, the development of tailored interventions may be an important avenue to improve response to treatment in this patient population.

The present study found a high rate of co-occurrence of HD in primary presenting LLD (24–30%), which was accompanied by an increased burden on executive functioning, disability, and depression treatment outcomes. These findings have important implications for intervention efforts. Hoarding symptoms are likely under-evaluated and thus may be overlooked and undertreated in clinical settings where LLD is identified as the primary diagnosis. Taken together with results indicating poorer depression treatment response in LLD+HD, our findings of elevated rates of hoarding in this population, and higher rates of cognitive and functional impairment underscore the need for increased screening of hoarding behaviors in LLD and the potential benefit of tailored interventions for this LLD+HD group. Future work examining the course of hoarding symptomatology in LLD (e.g., onset age of hoarding behaviors) may provide insights into the mechanisms associated with greater executive dysfunction and disability in this patient population.

Highlights:

What is the primary question addressed by this study?

What is the prevalence of Hoarding Disorder (HD) in Late Life Depression (LLD) and do hoarding symptoms in LLD contribute to greater executive dysfunction, disability and poorer depression treatment response?

What is the main finding of this study?

HD concurrence with LLD is common at 30.1%, which is higher than HD comorbidity with obsessive-compulsive disorder (≤18%) or HD prevalence in the general population of older adults (6%). Hoarding symptoms concurrent with LLD were associated with greater executive dysfunction, disability, and poorer response to depression treatment.

What is the meaning of the findings?

Individuals with HD and LLD may be at greater risk for cognitive and functional decline, thus underscoring the need for increased screening of hoarding symptoms in LLD and tailored interventions for individuals with both conditions.