Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Apr 27.
Published in final edited form as: Am J Geriatr Psychiatry. 2023 Nov 23;32(4):497–508. doi: 10.1016/j.jagp.2023.11.006

Relationship of Hoarding and Depression Symptoms in Older Adults

Sara Nutley 1,2,3,†, Binh K Nguyen 1,2,†, R Scott Mackin 4,5, Philip S Insel 4, Duygu Tosun 5,6, Meryl Butters 7, Paul Aisen 8,9, Rema Raman 8,9, Andrew J Saykin 10, Arthur W Toga 11, Clifford Jack 12, Michael W Weiner 4,5,6,13, Craig Nelson 4, Michelle Kassel 4,5, Maria Kryza-Lacombe 4,14, Joseph Eichenbaum 5,6, Rachel L Nosheny 4,6, Carol A Mathews 1,2,*
PMCID: PMC11055473  NIHMSID: NIHMS1985401  PMID: 38092621

Abstract

Hoarding disorder (HD) is a debilitating neuropsychiatric condition that affects 2-6% of the population and increases in incidence with age. Major depressive disorder (MDD) co-occurs with HD in approximately 50% of cases, and leads to increased functional impairment and disability. However, only one study to date has examined the rate and trajectory of hoarding symptoms in older individuals with a lifetime history of MDD, including those with current active depression (late life depression; LLD). We therefore sought to characterize this potentially distinct phenotype. We determined the incidence of HD in two separate cohorts of participants with LLD (n=73) or lifetime history of MDD (n=580) and examined the reliability and stability of hoarding symptoms using the Saving Inventory-Revised (SI-R) and Hoarding Rating Scale-Self Report (HRS), as well as the co-variance of hoarding and depression scores over time. HD was present in 12% to 33% of participants with MDD, with higher rates found in those with active depressive symptoms. Hoarding severity was stable across timepoints in both samples (all correlations>0.75), and fewer than 30% of participants in each sample experienced significant changes in severity between any two timepoints. Change in depression symptoms over time did not co-vary with change in hoarding symptoms. These findings indicate that hoarding is a more common comorbidity in LLD than previously suggested, and should be considered in screening and management of LLD. Future studies should further characterize the interaction of these conditions and their impact on outcomes, particularly functional impairment in this vulnerable population.

Keywords: late life depression (LLD), major depressive disorder (MDD), hoarding disorder (HD), Brain Health Registry (BHR), Saving Inventory-Revised (SI-R), Hoarding Rating Scale-Self Report (HRS), Hamilton Depression Rating Scale (HAM-D), Patient Health Questionnaire-9 (PHQ-9), reliable change, stability, trajectory analysis

Article Summary

This study examines the epidemiology of hoarding symptoms in two samples of older adults with depression and explores the stability of hoarding symptoms over time in relation to depression symptomatology. We found a substantial proportion - approximately 30% - of individuals with symptoms of current major depressive disorder had clinically significant hoarding symptoms, and that hoarding scores were relatively stable and did not co-vary with improvements in depression scores over time, suggesting other factors may contribute to the comorbidity of these psychiatric disorders.

Objective

Hoarding disorder (HD) is characterized by difficulty discarding possessions and accumulation of clutter that impairs use of living space and causes significant distress1. HD occurs in approximately 2% of the population2,3 and worsens with age, with incidence of up to 6% in older populations4. HD can lead to functional impairment and reduced quality of life5 and is highly comorbid with other medical and psychiatric conditions6,7. Co-occurring major depressive disorder (MDD) is particularly common, with reported rates in HD ranging from 14% to 54%, with most studies converging around 50%8–12.

Extant studies have focused primarily on rates of depression among individuals with known HD9,12,13; the reverse, that is, the rate of HD among individuals with MDD, is less well-studied14. HD continues to be under-recognized, under-diagnosed, and under-treated15, even among mental health professionals. Understanding the prevalence of this highly debilitating disorder, particularly among older individuals with other mental health conditions, is critical to improving identification of at-risk individuals.

While HD is most prevalent in older adulthood, MDD is common at all ages and occurs in up to 13% of older individuals (termed late life depression, or LLD), where it is associated with increased disability16,17. As HD increases in prevalence with age, the intersection between HD and LLD is of particular interest. Previous studies suggest that the overlap of depression and hoarding may be more functionally debilitating than hoarding or depression alone5,18,19. In a study examining hoarding participants with comorbid depression, individuals with both disorders reported greater hoarding severity and hoarding-related impairment, both of which were correlated with depression severity19. However, only one previous study has examined rates of HD among older individuals with MDD/LLD14. In this study, severe compulsive hoarding behaviors were found in 13% (n=7) of a sample of 52 participants with active LLD. Hoarding symptoms were associated with greater deficits in categorization, problem-solving, information processing, and verbal memory, suggesting a unique cognitive phenotype for comorbid LLD+HD14.

To date, most studies assessing the relationship between hoarding and depression have been cross-sectional. However, examining interactions between symptoms over time is an essential component for identifying and treating combined LLD+HD. Unlike depression in early and middle adulthood, LLD is often chronic and unremitting20, likely exacerbated by high comorbidity rates among older adults21. Though longitudinal investigations are limited, HD also appears chronic in nature. However, reports of symptom stability vary, with some reporting worsened severity with age22, and less frequently, symptoms that relapse and remit across the lifespan23. While psychiatric comorbidity, namely LLD, may influence hoarding symptom trajectory in older adulthood, the mechanisms underlying such possible relationships are not known, and the stability of combined symptomatology is not well-defined.

Thus, a primary aim of this study was to examine the prevalence of HD among older individuals with MDD using two independent datasets: individuals who met criteria for active depression participating in a larger study of LLD (UCSF sample)24 and individuals with a lifetime history of MDD enrolled in the Brain Health Registry (BHR; online sample)25,26. The former is composed of participants with confirmed LLD via structured clinical interview and self-reported hoarding data, whereas the latter is a large internet-based registry of participants with self-reported questionnaire data, including depression and hoarding history. We hypothesized that HD prevalence would be higher in those with MDD than in the general population (4-7%)2–4, based on the rate documented by Mackin et al. (2011)14. Additionally, we aimed to investigate the temporal stability of hoarding scores on two separate measures – the Saving Inventory, Revised (SI-R) and Hoarding Rating Scale, Self-Report (HRS) – and examine whether hoarding severity covaries with depression severity over time.

Methods

Participants:

The UCSF sample included 73 adults aged 65+ with MDD and active depressive symptoms at enrollment who also had available hoarding data (see Mackin et al. 2021 for details)24. The online sample consisted of 580 participants aged 65+ with a self-reported lifetime history of MDD and three or more assessments of hoarding and depression symptoms during a 30-month follow-up period. A subset of the online participants (n=113) also had active depressive symptoms at baseline assessment and were examined separately. Informed consent was obtained from all participants, and the study was approved by the University of California, San Francisco (UCSF) and University of Florida (UF) Institutional Review Boards.

Measures:

In the UCSF sample, hoarding was assessed using the Saving Inventory-Revised (SI-R)27, administered at 12, 24, and 30 months post-baseline (not all participants completed all timepoints). In the online sample, the Hoarding Rating Scale, Self-Report (HRS), was used to assess hoarding, administered every 6 months over 30 months (additional details in Supplementary Methods)28.

In the UCSF sample, depressive symptoms were assessed using the Hamilton Depression Rating Scale (HAM-D)29, administered at baseline and at 30-month follow-up. In the online sample, depressive symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9)30, administered at 6-month intervals over 30 months.

UCSF participants completed in-person psychiatric assessment using the Structured Clinical Interview for DSM Disorders31. Participants in the online sample were asked to self-report lifetime history (yes/no) of psychiatric conditions. The accuracy of psychiatric self-report data in the online sample as compared to clinical interview ranges from 0.74 for MDD to 0.97 for bipolar disorder32. HD diagnosis accuracy in the online sample was 0.97 using HRS cutoff scores33.

HD and LLD definitions:

To determine HD case/control status, we averaged scores for the primary measure of hoarding in each sample across all timepoints. In the UCSF sample, an average SI-R score of ≥33 indicated HD, a cutoff score determined to be the most appropriate in older populations34. If a participant had only one SI-R score, that score was used to assign HD/non-HD status. In the online sample, an HRS cutoff score of ≥ 14 was used to indicate HD28,33. All UCSF participants met criteria for LLD, defined by consensus diagnosis and a HAM-D score of ≥15 at baseline. In the online sample, a lifetime history of MDD and a PHQ-9 score of ≥10 at baseline35 indicated LLD (n=113/580).

Analyses:

The primary aim of this investigation was to examine the prevalence of HD among older individuals with LLD. After identifying the proportion of participants affected by hoarding in each sample, demographic characteristics were compared between hoarding groups using Pearson’s chi-square, Fisher’s exact, and independent sample t-tests, as appropriate. In addition, rates of psychiatric comorbidities (at baseline) were compared between groups in the online sample. A p-value of <0.05 was considered statistically significant.

Stability Analysis and Reliable Change:

Our second aim was to investigate the temporal stability of hoarding symptoms measured via the SI-R and HRS. In each sample, independent sample t-tests were used to compare depression (HAM-D/PHQ-9) and hoarding (SI-R, HRS) scores among those with and without hoarding at each timepoint. Further, bivariate Pearson correlations were calculated to assess stability of SI-R scores over time for participants in the UCSF sample with complete data at 2+ timepoints (N=41), and of HRS scores in the online sample (N=580). Paired samples t-tests were conducted between timepoints to identify statistically significant change over time. Finally, reliable change indices (RCI) were calculated for the SI-R and HRS36 to identify meaningful change in hoarding severity across timepoints (details in Supplemental Methods).

Trajectory Analysis:

Our third and final aim was to examine whether hoarding severity covaries with depression severity over time. We determined the longitudinal trajectory of HRS and PHQ-9 mean scores in the online sample only (due to limited sample timepoints in the UCSF sample), and used random intercept cross-lagged panel models (RI-CLPMs) to assess the relationships between depressive and hoarding symptoms across time. Parameters were estimated using full information maximum likelihood. The RI-CPLM was first estimated without constraints. Autoregressive and cross-lagged effects, residual variances and covariances, and grand means were then constrained to be time-invariant. Models were compared using a chi-square difference test, information criteria (i.e., Akaike information criteria [AIC], Bayesian information criteria [BIC], and recommended fit indices: the root mean square error of approximation (RMSEA), comparative fit index (CFI), and Tucker-Lewis index (TLI).

Results

Demographic and clinical characteristics:

Both the UCSF sample (n=73) and the online sample (N=580) were predominately female, white, highly educated, and of older age (Table 1). Twenty-one participants in the UCSF sample (28.8%) had a mean SI-R score ≥33 and were classified in the HD group. Age, gender, race, and education did not significantly differ between HD and non-HD groups (Table 1). There were no differences on demographic factors or depression measures between those who had hoarding data at 1 timepoint vs. those who had hoarding data at 2+ timepoints (data not shown). Seventy-five participants in the online sample with self-reported lifetime history of MDD (12.9%) were classified in the HD group (mean HRS total score ≥14). Age, gender, race, and education did not significantly differ between HD and non-HD groups. One hundred and thirteen participants (19.5%) endorsed active depressive symptoms or LLD (i.e., PHQ total score ≥10) at baseline. Among these participants, 34 endorsed hoarding behavior (30.1%). Age of hoarding symptom onset was comparable across samples, with a mean of 34.8 (n=13, SD=21.7) and 38.5 years (n=67, SD=20.1) in the UCSF and online samples, respectively.

Table 1.

Demographic characteristics of both samples and psychiatric comorbidities in the online sample.

UCSF Sample
HD (SI-R ≥ 33)
n=21 (28.8%)		 Non-HD (SI-R < 33)
n=52 (71.2%)		 Total
n=73		 test statistic, p-value
Mean (SD) age at baseline (yrs) 69.6 (3.9), range = 65-78 70.4 (5.0)
range = 65-88		 70.2 (4.7)
range = 65-88		 t(71) = 0.63, p = 0.53
Female (gender) 12
(57.1%)		 36
(69.2%)		 48
(65.8%)		 χ2(1) = 0.97, p = 0.32
White (race) 18
(85.7%)		 45
(86.5%)		 63
(86.3%)		 χ2(1) = 0.01, p = 0.93
Mean (SD) education (yrs) 16.1 (2.0)
range = 14-20		 16.6 (2.1) range = 12-20 16.5 (2.1)
range = 12-20		 t(71) = 0.92, p = 0.36
Online Sample
HD (HRS ≥ 14)
n=75 (12.9%)		 Non-HD (HRS < 14)
n=505 (87.1%)		 Total
n=580		 test statistic, p-value
Mean (SD) age at baseline (yrs) 69.5 (3.4) 69.4 (3.9) 69.4 (3.8) t(104) = −0.4 p = 0.69
Female (gender) 62 (82.7%) 407 (80.6%) 469 (80.9%) X2 (1) = 0.2, p = 0.67
White (race) 68 (90.7%) 486 (96.2%) 554 (95.5%) X2 (1) = 4.7 p = 0.03
Mean (SD) education (yrs) 16.3 (2.2) 16.8 (2.4) 16.7 (2.4) t(102) = 1.9 p = 0.06
Generalized Anxiety Disorder 30 (40.0%) 186 (36.8%) 216 (37.2%) X2 (1) = 0.28, p = 0.60
Panic Disorder 12 (16.0%) 60 (11.9%) 72 (12.4%) X2 (1) = 1.02, p = 0.31
Post-Traumatic Stress disorder 20 (26.7%) 92 (18.2%) 112 (19.3%) X2 (1) = 2.99, p = 0.08
Obsessive Compulsive Disorder 11 (14.7%) 24 (4.8%) 35 (6.0%) X2 (1) = 11.32, p=0.001
Eating Disorder 13 (17.3%) 42 (8.3%) 55 (9.5%) X2 (1) = 6.18, p = 0.013
Substance Use Disorder 19 (25.3%) 98 (19.4%) 117 (20.2%) X2 (1) = 1.42, p = 0.23
Attention Deficit/Hyperactivity Disorder 21 (28.0%) 48 (9.5%) 69 (11.9%) X2 (1) = 21.31, p<0.001
Online Sample with Current Depression (LLD) (n=113)
HD (HRS ≥ 14)
n=34 (30.1%)		 Non-HD (HRS < 14)
n=79 (69.9%)		 Total
n=113		 test statistic, p-value
Mean (SD) age at baseline (yrs) 69.6 (3.4) 69.4 (3.8) 69.5 (3.7) t(69) = −0.2 p = 0.81
Female (gender) 27 (79.4%) 59 (74.7%) 89 (76.1%) X2 (1) =0.29, p = 0.59
White (race) 32 (94.1%) 74 (93.7%) 106 (93.8%) X2 (1) = 0.008, p = 0.93
Mean (SD) education (yrs) 16.2 (2.5) 16.0 (2.2) 16.1 (2.4) t(70) = −0.4 p = 0.72
Generalized Anxiety Disorder 15 (44.1%) 48 (60.8%) 63 (55.8%) X2 (1) = 2.67, p = 0.10
Panic Disorder 5 (14.7%) 15 (19.0%) 20 (17.7%) X2 (1) = 0.30, p = 0.58
Post-Traumatic Stress disorder 12 (35.3%) 21 (26.6%) 33 (29.2%) X2 (1) = 0.87, p = 0.35
Obsessive Compulsive Disorder (n<5) (n<5) (n<10) p = 0.195
Eating Disorder 6 (17.6%) 10 (12.7%) 16 (14.2%) X2 (1) = 0.49, p = 0.49
Substance Use Disorder 9 (26.5%) 14 (17.7%) 23 (20.4%) X2 (1) = 1.12, p = 0.29
Attention Deficit/Hyperactivity Disorder 12 (35.3%) 8 (10.1%) 20 (17.7) X2 (1) = 10.34, p = 0.001
Open in a new tab

Test statistics and corresponding p-values from independent sample t-tests, Pearson’s chi-square test, and Fisher’s exact test, as appropriate.

SD: standard deviation

Clinical comorbidities in the samples:

Clinical characteristics of the UCSF cohort have been previously reported24. In the online sample, other than self-reported lifetime history of MDD, which by definition was endorsed by all participants, generalized anxiety disorder (GAD) was the most commonly reported psychiatric diagnosis (37.2%), followed by substance use disorder (SUD; 20.2%) and post-traumatic stress disorder (19.3%) (Table 1). Compared to those without HD, the HD group had higher rates of self-reported OCD (14.7% vs. 4.8%, X2 (1) = 11.32, p=0.001), ADHD (28.0% vs. 9.5%, X2 (1) = 21.31, p<0.001), and eating disorders (17.3% vs. 8.3%, X2 (1) = 6.18, p=0.01). Psychiatric comorbidity rates were similar in the subset of participants with current depressive symptoms (LLD), though only ADHD was significantly associated with hoarding (HD: 35.3% vs. non-HD: 10.1%, X2 (1) = 10.34, p=0.001).

Depression severity across time:

Hoarding and depression symptom severity in HD and non-HD participants over time are depicted in Figures 1–3 (hoarding subscales in Supplemental Figures 1–3). While depression severity decreased over time in the UCSF sample (Figure 1), we observed no significant difference in depression severity among participants with and without HD at baseline or 30-month follow-up (Baseline: 19.1 (3.6) vs. 17.6 (2.0), t(25) = −1.84, p=0.08; 30 month: 11.1 (6.2) vs. 8.3 (5.6), t(65) = −1.75, p=0.09; Supplemental Table 1). In contrast, in the full online sample, depression severity was higher among participants with HD relative to those without at all timepoints, with mean total scores in both groups remaining relatively stable over time (Figure 2 and Supplemental Table 2). Among online participants with LLD, depression severity decreased over time, and differed between HD and non-HD groups only at the 6 and 12-month timepoints (6 month: HD mean (SD) = 14.8 (5.7) vs non-HD mean (SD) = 10.7 (5.1), t(34)=−7.2, p=0.005; 12 month: HD mean (SD) = 14.6 (6.1) vs non-HD mean (SD) = 10.3 (5.2), t(32)=−3.0, p=0.005; Figure 3).

Figure 1.

Figure 1.

Open in a new tab

Mean scores in the UCSF sample (n=73) on the SI-R (hoarding) and HAM- D (depression) measures, along with individual trajectories of all participants. Red = HD/LLD. Blue = LLD. The vertical lines represent standard deviation (SD). The blue horizontal lines in each graph represent clinically significant cutoffs on each measure (33 for the SI-R and 15 for the HAM-D). For SI-R total score, the two groups significantly differed (p<0.05) from one another at all timepoints. For HAM-D total score, the two groups did not significantly differ from one another at either timepoint (test statistics and p-values provided in Supplemental Table 1).

Figure 3.

Figure 3.

Open in a new tab

Mean scores in the online sample with active depression (PHQ-9 ≥10 at baseline, n=113) on the HRS-SR (hoarding) and PHQ-9 (depression) measures. Red = HD/LLD. Blue = LLD. The vertical lines represent standard deviation (SD). The blue horizontal lines represent clinically significant cutoffs on each measure (14 for the HRS-SR and 10 for the PHQ-9). Individual trajectories are not shown due to the larger sample size and greater number of timepoints in the online sample. For HRS-SR total score, the two groups significantly differed (p<0.05) from one another at all timepoints. For PHQ-9 total score, the two groups significantly differed from one another (p<0.05) at 6 months and 12 months only (test statistics and p-values provided in Supplemental Table 2).

Figure 2.

Figure 2.

Open in a new tab

Mean scores in the online sample (n=580) on the HRS-SR (hoarding) and PHQ-9 (depression) measures. Red = HD. Blue = Non-HD. The vertical lines represent standard deviation (SD). The blue horizontal lines represent clinically significant cutoffs on each measure (14 for the HRS-SR and 10 for the PHQ-9). Individual trajectories are not shown due to the larger sample size and greater number of timepoints in the online sample. For both HRS-SR total score and PHQ-9 total score, the two groups significantly differed (p<0.05) from one another at all timepoints (test statistics and p-values provided in Supplemental Table 2).

Hoarding severity across time:

In the UCSF sample, 41/73 participants had SI-R scores at multiple timepoints, 35 (85.4%) with scores at two timepoints and 6 (14.6%) with scores at three timepoints, for a total of 53 timepoint pairs. Bivariate Pearson correlation coefficients between SI-R total scores at 12, 24 and 30 months ranged between 0.73 and 0.91 (Supplemental Figure 4). SI-R subscales were also strongly correlated across time (Supplemental Figures 5–7). Paired-samples t-tests between SI-R total scores were non-significant (t(38) = −0.274, p=0.79 for months 24 and 30, t(6) = 0.83, p=0.44 for months 12 and 30, and t(6) = −0.05, p=0.96 for months 12 and 24).

In the online sample, bivariate Pearson correlations between HRS total scores were also high, especially between consecutive timepoint pairs (all correlation coefficients between 0.75 and 0.83, Supplemental Figure 8). The magnitude of the correlations between HRS subscore measures across time were also moderate to large in size (all correlations between 0.55 and 0.76) (Supplemental Figures 9–11, only 0, 12, 24, and 30 months depicted). A similar pattern was observed among online participants with LLD, wherein correlation coefficients between HRS total scores ranged between 0.67 and 0.90 across timepoint pairs (Supplemental Figures 12–15, only 0, 12, 24, and 30 months depicted).

Reliable Change:

For the SI-R, a previously reported Cronbach’s alpha value of 0.9637 and the standard deviation of SI-R total scores in the full UCSF sample at baseline (14.81) were used to calculate an RCI of 8.21. Between months 24 and 30 (the timepoint pair with the most data), 10 of 39 participants (25.6%; 3/11 or 27.3% with HD and 7/28 or 25%,without HD) met criteria for reliable change: 5 were classified as RCI+ (SI-R increased more than 8.21 points from 24 to 30 months), and 5 as RCI− (SI-R decreased more than 8.21 points from 24 to 30 months). Four participants (10.3% of the 39) had an SI-R score change that exceeded the RCI and also caused change in HD classification, with two moving from HD to non-HD and two from non-HD to HD.

For the HRS, a previously reported Cronbach’s alpha value of 0.8738 and the standard deviation of HRS scores in the online sample at baseline (6.21) were used to calculate an HRS RCI of 6.2. Sixty-nine of the 580 individuals with MDD in the online sample (12%) had an HRS score change that exceeded the RCI between any timepoint pair: 29 were classified as RCI+ and 40 were classified as RCI−. However, only 30 participants (5.2%) had an absolute change between consecutive time points greater than or equal to 6.2 that also caused change in HD classification (N=13 or 17.3% with HD; N=17 or 3.4% without HD). In the online sample of those with LLD, 31 of the 113 participants (27.4%) had an HRS score change that exceeded the RCI (RCI+ n=14, RCI− n=17), though only 12 participants (10.6%) exceeded the RCI and also changed classification.

Trajectory and Covariation of Hoarding and Depression:

Model fit statistics for the unconstrained and constrained RI-CLPMs in the online sample indicated exceptional model fit (Unconstrained RMSEA: 0.031 [0.014, 0.046], CLI: 0.995, TLI: 0.991; Constrained RMSEA: 0.026 [0.011, 0.037], CLI: 0.993, TLI: 0.994). As the chi-square difference test indicated no change in model fit following the addition of equality constaints (Δχ2(38) = 45.6, p = 0.19; Unconstrained AIC: 9768, BIC: 9999; Constrained AIC: 9738, BIC: 9803), the more parsimonious model (i.e., the constrained RI-CLPM) was selected as the final model (Figure 4). As indicated by the significant between-person association between depressive and hoarding symptoms (correlation coefficient, r=0.43, z = 10.64, p<0.001), participants with more severe depressive symptoms over the course of the follow-up period were also more likely to endorse hoarding symptoms. In terms of within-person effects, significant autoregressive paths for hoarding suggested a carry-over effect wherein higher than expected hoarding symptomatology at the previous time point was a significant predictor of higher than expected hoarding symptomatology at the current time point (unstandardized β=0.13, z = 3.74, p<0.001). However, autoregressive paths for depression were not significant (unstandardized β=0.002, z = 0.06, p = 0.952). Small, albeit signifcant, correlations between hoarding and depressive symptoms at the within-person level indicated that higher than expected depressive symptoms at a given time point were associated with higher than expected hoarding symptoms at the same time point (residual covariance=0.04, z = 6.53, p<0.001). However, there were no significant cross-lagged paths between hoarding and depression, indicating that higher than expected depression severity did not predict higher than expected hoarding severity six months later, and vice versa (Hoarding to Depression: unstandardized β=0.01, z = 0.306, p=0.760; Depression to Hoarding: unstandardized β=0.01, z = 0.192, p = 0.848; Supplemental Figure 16). Although not formally assessed in the UCSF sample, Figure 1 demonstrates that hoarding remained stable while depression decreased at follow-up, suggesting no substantial covariation.

Figure 4.

Figure 4.

Open in a new tab

Path diagram displaying significant covariances and regression effects for the constrained RI-CLPM assessing the bidirectional relationships between hoarding depressive symptoms over time. Between-person effects for hoarding (HRS) and depressive symptoms (PHQ) are labeled using the subscript ‘bp,’ while within-person effects are labeled using the subscript ‘wp,’ followed by the respective time point in months. Insignificant paths are depicted using a dashed line. Significant paths (at the 0.05 level) are depicted using a solid line, and corresponding path coefficients (or correlations/covariances) are marked with an *.

Significant effects include: a between-person association between depressive and hoarding symptoms (r=0.43, z = 10.64, p<0.001), within-person autoregressive paths for hoarding symptoms (unstandardized β=0.13, z = 3.74, p<0.001), and within-person correlations between hoarding and depressive symptoms (residual covariance=0.04, z = 6.53, p<0.001).

Conclusions

The results of this study describe the rates of HD among older individuals with MDD, examine the relative stability and covariance of hoarding and depressive symptoms across time, and establish reliable change indices for the SI-R and HRS across two separate samples. Rates of HD in these samples (12%-33%) were 2-5 times higher than the highest reported population prevalence rates in this age group4. While the rate of HD among individuals with a lifetime history of MDD but not necessarily active depressive symptoms in the online sample was comparable to the rates found previously by Mackin et al (13.3%), the prevalence more than doubled when HD was assessed in those with active depression symptoms at baseline (LLD), where approximately 30% met the cutoff for clinically significant HD – rates that coincide between samples despite considerable variation in sampling and measurement. These findings suggest the potential utility of screening older individuals with a history of MDD, and in particular, those with active symptoms of depression, for the presence of hoarding symptoms. Given the elevated rates of functional impairment and reduced quality of life reported by those with combined hoarding and depressive symptoms14,19, clarification of hoarding comorbidity among older adults with LLD may be essential for identifying appropriate course of treatment and reducing related impairment and disability. The high rates of comorbidity between HD and MDD, as well as recent findings from network analysis suggesting primary relationships between these disorders7 also suggests that further examination into the etiologies of these overlapping phenotypes is warranted.

Although we observed no differences in depression symptom severity among older adults with and without HD when limiting analysis to those with current active depression (LLD) in either sample, we did find that individuals in the online sample with lifetime MDD and co-occurring HD reported more severe depressive symptoms at every timepoint than those without HD. We also found significantly higher lifetime history rates of OCD, eating disorders, and ADHD in those with HD+MDD compared to those with MDD alone. Taken together, these findings suggest that the presence of hoarding symptoms among older individuals with depression may suggest a higher psychiatric burden in general, which in turn may correspond to higher rates of disability and impairment in this already vulnerable population.

Although the cross lagged model in the online sample provided some evidence for slight worsening of hoarding symptoms across time, overall, hoarding symptom severity remained relatively stable for most participants in both samples, despite the use of different assessments. In the UCSF sample, reliable change analyses suggested that approximately one fourth of participants, regardless of hoarding status, showed significant changes in scores on the SI-R, while only 10.3% exceeded the reliable change index and also crossed clinical categories. Fewer than 20% of those with hoarding in the online sample had reliable change at one or more timepoints on the HRS and also crossed clinical categories, compared to 3.4% of those without hoarding, implying that the majority of the sample did not exhibit meaningful change on the HRS.

Finally, for the majority of participants with HD, hoarding and depression did not covary across time. This suggests that although hoarding and depression commonly co-occur with one another, they are not necessarily co-dependent. Changes in hoarding scores were not solely attributable or related to changes in depression scores, and improvements in depression over time (which was seen in both LLD groups) did not lead to similar changes in hoarding symptom severity.

Limitations

Although this study has several strengths, including the use of two independent datasets and the availability of longitudinal data, it also has limitations. First, neither the UCSF nor the online cohorts were epidemiological samples, and thus we cannot make definitive statements regarding population prevalences of LDD+HD based on our results. Second, the samples varied in size, and the criteria for study participation and inclusion in these secondary analyses also differed. Diagnoses of HD were assigned using cutoff criteria on self-report hoarding symptom scales, and it was not possible to confirm diagnoses with a clinical interview. Psychiatric diagnoses in the online sample were also self-reported and may be subject to bias. Moreover, both samples were predominately female, white, and highly educated. As epidemiological data that describe variation in the prevalence of hoarding behavior by gender and race/ethnicity are limited4,39, it is difficult to estimate the potential impact of sample demographic characteristics on observed prevalence. Future work would benefit from larger and more diverse samples in order to improve the generalizability of these results to the general population of individuals with LLD.

Despite differences in characteristics and ascertainment, the results were remarkably consistent between the two samples, and provide evidence for higher rates of clinically significant hoarding symptoms in older adults with depression than previously suggested, emphasizing the need to screen for and address hoarding symptoms in this population, as the presence of both conditions may lead to worse outcomes. Similarly, clinicians should assess for and treat symptoms of depression among individuals with hoarding symptoms. Pharmacotherapy and psychotherapy have efficacy for the treatment of both depression and hoarding, and there is some overlap in agents and approaches (e.g., some antidepressants also have some efficacy in treating hoarding)40. Future work should continue to explore the complex intersection between hoarding and LLD in order to develop better treatments, determine risk factors for the development of hoarding among those with depression, and investigate variables that contribute to changes in hoarding symptoms and severity.

Supplementary Material

1

Acknowledgements

This work was funded by NIH grant MH117114 and by the University of Florida Center for OCD, Anxiety, and Related Disorders (COARD).

The UCSF sample was supported by NIMH grant R01098062 and research support by the Ray and Dagmar Dolby Family Fund.

The online Brain Health Registry sample was supported by NIA grant R33AG062867.

Footnotes

Disclosures/Conflicts of Interest

The authors report no conflicts with any product mentioned or concept discussed in this article.

Data Availability Statement

The data that support the findings of this study are available from the Brain Health Registry (PI Michael Weiner) and/or the UCSF sample (PI Scott Mackin) following submission of a Data Sharing Request Form. Additional information is available at www.brainhealthregistry.org/for-investigators/de-identified-data-sharing for the Brain Health Registry data.

References

  • 1.Association AP. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Arlington, VA: 2013. [Google Scholar]
  • 2.Samuels JF, Bienvenu OJ, Grados MA, et al. Prevalence and correlates of hoarding behavior in a community-based sample. Behaviour Research and Therapy. 2008-July-01 2008;46(7):836–844. doi: 10.1016/j.brat.2008.04.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Postlethwaite A, Kellett S, Mataix-Cols D. Prevalence of Hoarding Disorder: A systematic review and meta-analysis. Journal of Affective Disorders. 2019-September-01 2019;256:309–316. doi: 10.1016/j.jad.2019.06.004 [DOI] [PubMed] [Google Scholar]
  • 4.Cath DC, Nizar K, Boomsma D, Mathews CA. Age-Specific Prevalence of Hoarding and Obsessive Compulsive Disorder: A Population-Based Study. The American Journal of Geriatric Psychiatry. 2017-March-01 2017;25(3):245–255. doi: 10.1016/j.jagp.2016.11.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Nutley SK, Read M, Eichenbaum J, et al. Health-related quality of life in hoarding: A comparison to chronic conditions with high disease burden. J Psychiatr Res. Mar 2 2022;149:68–75. doi: 10.1016/j.jpsychires.2022.02.035 [DOI] [PubMed] [Google Scholar]
  • 6.Nutley SK, Camacho MR, Eichenbaum J, et al. Hoarding disorder is associated with self-reported cardiovascular / metabolic dysfunction, chronic pain, and sleep apnea. J Psychiatr Res. Feb 2021;134:15–21. doi: 10.1016/j.jpsychires.2020.12.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sordo Vieira L, Guastello A, Nguyen B, et al. Identifying psychiatric and neurological comorbidities associated with hoarding disorder through network analysis. J Psychiatr Res. Dec 2022;156:16–24. doi: 10.1016/j.jpsychires.2022.09.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ayers CR, Saxena S, Golshan S, Wetherell JL. Age at onset and clinical features of late life compulsive hoarding. International Journal of Geriatric Psychiatry. 2010-February-01 2010;25(2):142–149. doi: 10.1002/gps.2310 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Frost RO, Steketee G, Tolin DF. Comorbidity in hoarding disorder. Depression and Anxiety. 2011-October-03 2011;28(10):876–884. doi: 10.1002/da.20861 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Diefenbach GJ, Dimauro J, Frost R, Steketee G, Tolin DF. Characteristics of Hoarding in Older Adults. The American Journal of Geriatric Psychiatry. 2013-October-01 2013;21(10):1043–1047. doi: 10.1016/j.jagp.2013.01.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Roane DM, Landers A, Sherratt J, Wilson GS. Hoarding in the elderly: a critical review of the recent literature. International Psychogeriatrics. 2017-July-01 2017;29(7):1077–1084. doi: 10.1017/s1041610216002465 [DOI] [PubMed] [Google Scholar]
  • 12.Archer CA, Moran K, Garza K, et al. Relationship between symptom severity, psychiatric comorbidity, social/occupational impairment, and suicidality in hoarding disorder. Journal of Obsessive-Compulsive and Related Disorders; 2019. p. 158–164. [Google Scholar]
  • 13.Sordo Vieira L, Nguyen B, Nutley S, Mackin RS, Mathews CA. Comorbidity Patterns of Hoarding Disorder. Biological Psychiatry. 2021-May-01 2021;89(9):S133–S134. doi: 10.1016/j.biopsych.2021.02.344 [DOI] [Google Scholar]
  • 14.Mackin RS, Arean PA, Delucchi KL, Mathews CA. Cognitive functioning in individuals with severe compulsive hoarding behaviors and late life depression. Int J Geriatr Psychiatry. Mar 2011;26(3):314–21. doi: 10.1002/gps.2531 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Mataix-Cols D, Fernández De La Cruz L Hoarding disorder has finally arrived, but many challenges lie ahead. World Psychiatry. 2018-June-01 2018;17(2):224–225. doi: 10.1002/wps.20531 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Van Damme A, Declercq T, Lemey L, Tandt H, Petrovic M. Late-life depression: issues for the general practitioner. International Journal of General Medicine. 2018-March-01 2018;Volume 11:113–120. doi: 10.2147/ijgm.s154876 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Rhodes E, Insel PS, Butters MA, et al. The Impact of Amyloid Burden and APOE on Rates of Cognitive Impairment in Late Life Depression. J Alzheimers Dis. 2021;80(3):991–1002. doi: 10.3233/JAD-201089 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Ong C, Pang S, Sagayadevan V, Chong SA, Subramaniam M. Functioning and quality of life in hoarding: A systematic review. J Anxiety Disord. May 2015;32:17–30. doi: 10.1016/j.janxdis.2014.12.003 [DOI] [PubMed] [Google Scholar]
  • 19.Ohlhausen A, Dozier M, Davidson E, Ayers C. A Comparison of Hoarding Disorder With and Without Comorbid Major Depressive Disorder. Innovation in Aging; 2018. p. 683. [Google Scholar]
  • 20.de la Torre-Luque A, de la Fuente J, Sanchez-Niubo A, et al. Stability of clinically relevant depression symptoms in old-age across 11 cohorts: a multi-state study. Acta Psychiatr Scand. Dec 2019;140(6):541–551. doi: 10.1111/acps.13107 [DOI] [PubMed] [Google Scholar]
  • 21.Ismail Z, Fischer C, McCall WV. What characterizes late-life depression? Psychiatr Clin North Am. Dec 2013;36(4):483–96. doi: 10.1016/j.psc.2013.08.010 [DOI] [PubMed] [Google Scholar]
  • 22.Dozier ME, Porter B, Ayers CR. Age of onset and progression of hoarding symptoms in older adults with hoarding disorder. Aging Ment Health. Jul 2016;20(7):736–42. doi: 10.1080/13607863.2015.1033684 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Tolin DF, Meunier SA, Frost RO, Steketee G. Course of compulsive hoarding and its relationship to life events. Depress Anxiety. Sep 2010;27(9):829–38. doi: 10.1002/da.20684 [DOI] [PubMed] [Google Scholar]
  • 24.Mackin RS, Insel PS, Landau S, et al. Late-Life Depression Is Associated With Reduced Cortical Amyloid Burden: Findings From the Alzheimer’s Disease Neuroimaging Initiative Depression Project. Biol Psychiatry. Apr 15 2021;89(8):757–765. doi: 10.1016/j.biopsych.2020.06.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Weiner MW, Nosheny R, Camacho M, et al. The Brain Health Registry: An internet‐based platform for recruitment, assessment, and longitudinal monitoring of participants for neuroscience studies. Alzheimer’s & Dementia. 2018-August-01 2018;14(8):1063–1076. doi: 10.1016/j.jalz.2018.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Weiner MW, Aaronson A, Eichenbaum J, et al. Brain health registry updates: An online longitudinal neuroscience platform. Alzheimers Dement. Mar 25 2023;doi: 10.1002/alz.13077 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Frost RO, Steketee G, Grisham J. Measurement of compulsive hoarding: saving inventory-revised. Behav Res Ther. Oct 2004;42(10):1163–82. doi: 10.1016/j.brat.2003.07.006 [DOI] [PubMed] [Google Scholar]
  • 28.Tolin DF, Frost RO, Steketee G. A brief interview for assessing compulsive hoarding: The Hoarding Rating Scale-Interview. Psychiatry Research. 2010-June-01 2010;178(1):147–152. doi: 10.1016/j.psychres.2009.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hamilton M. A Rating Scale for Depression. Journal of Neurology, Neurosurgery & Psychiatry. 1960-February-01 1960;23(1):56–62. doi: 10.1136/jnnp.23.1.56 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kroenke K, Spitzer RL, Williams JBW. The PHQ-9. Journal of General Internal Medicine. 2001-September-01 2001;16(9):606–613. doi: 10.1046/j.1525-1497.2001.016009606.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.First MB, Williams JW, Karg RS, Spitzer RL. Structured Clinical Interview for DSM-5—Research Version (SCID-5 for DSM-5, Research Version; SCID-5-RV). Arlington, VA: American Psychiatric Association; 2015. [Google Scholar]
  • 32.Sordo Vieira L, Nguyen B, Nutley SK, et al. Self-reporting of psychiatric illness in an online patient registry is a good indicator of the existence of psychiatric illness. J Psychiatr Res. Apr 15 2022;151:34–41. doi: 10.1016/j.jpsychires.2022.03.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Nutley SK, Bertolace L, Vieira LS, et al. Internet-based hoarding assessment: The reliability and predictive validity of the internet-based Hoarding Rating Scale, Self-Report. Psychiatry Res. Dec 2020;294:113505. doi: 10.1016/j.psychres.2020.113505 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Kellman-Mcfarlane K, Stewart B, Woody S, et al. Saving inventory – Revised: Psychometric performance across the lifespan. Journal of Affective Disorders. 2019-June-01 2019;252:358–364. doi: 10.1016/j.jad.2019.04.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Levis B, Benedetti A, Thombs BD, Collaboration DESD. Accuracy of Patient Health Questionnaire-9 (PHQ-9) for screening to detect major depression: individual participant data meta-analysis. BMJ. Apr 9 2019;365:l1476. doi: 10.1136/bmj.11476 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Jacobson NS, Truax P. Clinical significance: a statistical approach to defining meaningful change in psychotherapy research. J Consult Clin Psychol. Feb 1991;59(1):12–9. doi: 10.1037//0022-006x.59.1.12 [DOI] [PubMed] [Google Scholar]
  • 37.Ayers CR, Dozier ME, Mayes TL. Psychometric Evaluation of the Saving Inventory-Revised in Older Adults. Clinical Gerontologist. 2017-May-27 2017;40(3):191–196. doi: 10.1080/07317115.2016.1267056 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Tolin DF, Gilliam CM, Davis E, et al. Psychometric properties of the Hoarding Rating Scale-Interview. Journal of Obsessive-Compulsive and Related Disorders. 2018-January-01 2018;16:76–80. doi: 10.1016/j.jocrd.2018.01.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Dozier ME, Ayers CR. The Etiology of Hoarding Disorder: A Review. Psychopathology. 2017;50(5):291–296. doi: 10.1159/000479235 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Saxena S, Sumner J. Venlafaxine extended-release treatment of hoarding disorder. Int Clin Psychopharmacol. Sep 2014;29(5):266–73. doi: 10.1097/YIC.0000000000000036 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1985401-supplement-1.pdf (2.8MB, pdf)

Data Availability Statement

The data that support the findings of this study are available from the Brain Health Registry (PI Michael Weiner) and/or the UCSF sample (PI Scott Mackin) following submission of a Data Sharing Request Form. Additional information is available at www.brainhealthregistry.org/for-investigators/de-identified-data-sharing for the Brain Health Registry data.

RESOURCES