Abstract
Background:
Despite consistent support for smaller total hippocampal volumes in late-life depression (LLD), there is less consensus on LLD’s relationship with hippocampal subregions and subfields. This is particularly germane for interactive effects of aging in context of depression, as individuals with LLD may exhibit accelerated aging processes.
Methods:
We examined 260 depressed (current and recently remitted) and 140 non-depressed older adults who completed 3T MRI and extracted hippocampal subregion and subfield volumes. Primary analyses investigated effects of LLD diagnosis, depression status, and age of onset on hippocampal volumes. Secondary analyses evaluated interactive effects of age and these diagnostic groupings on hippocampal volumes.
Results:
Compared to controls, LLD exhibited smaller total hippocampal volumes, with smaller volumes of the bilateral hippocampal head and the right hemisphere body, tail, CA1, and molecular layer. When comparing current and remitted LLD, individuals with current LLD exhibited smaller right total hippocampus, head, body, and molecular layer volumes compared to controls. They also exhibited smaller CA1, CA2, and CA4/dentate gyrus volumes compared to both control and remitted LLD groups. No regions differed between control and remitted LLD groups. We did not observe significant group differences in any volume measure between LLD individuals with an early- or later-life depression onset. There was no significant age by diagnostic group interactions on hippocampal volumes across any analyses.
Conclusions:
This cross-sectional work supports diagnostic, but not accelerated aging, effects of depression on hippocampus volumes. In contrast to recently remitted LLD, currently depressed individuals exhibited smaller volumes primarily in the right hippocampus.
Keywords: Depressive Disorder, hippocampus, aging, depression, geriatric
INTRODUCTION
Late-life depression (LLD) is a debilitating neuropsychiatric condition with poor health outcomes (1, 2), including high rates of medical morbidity, increased mortality, cognitive impairment and dementia. Given the relationship between LLD and cognitive impairment, studies in LLD have examined structural brain differences in regions influencing cognitive performance (2), including the hippocampus. Smaller hippocampal volumes are reported in both LLD and adult major depressive disorder (MDD) across the lifespan (3, 4), and smaller hippocampal volumes may be seen even with subclinical depressive symptoms (5). Conversely, larger hippocampal volumes may be a marker of positive response to antidepressant treatment (6). In LLD (7, 8), smaller hippocampal volumes are associated with incident dementia (9), although this finding is not necessarily specific to Alzheimer’s disease (AD) pathology (10). Most past work examines the hippocampus as a solitary unit rather than examining functionally relevant subregions and subfields (11, 12).
The hippocampus can be divided into three subregions along the anterior-posterior axis (head, body, and tail), each with distinct anatomical connections and functional roles (13, 14). The anterior head processes emotional, social, and semantic stimuli, while the posterior tail processes spatial and visual information (13–15). The hippocampus also includes functionally distinct subfields, including cornu ammonis (CA) regions 1–4, dentate gyrus, and subiculum. These subfields have distinct projections and are differentially related to memory and other cognitive processes (12). In contrast to relatively consistent findings of smaller total hippocampal volumes in depressed populations (3, 4, 7, 8), there is less consensus on which subregions and subfields are most affected in adult MDD or LLD. Past analyses support smaller hippocampal tail volumes bilaterally in adult MDD, although right hippocampal head has also been implicated (16, 17). Similarly, while some studies did not observe subfield volume differences in adult MDD, a meta-analysis associated MDD with decreased volumes in left hemisphere CA3 and CA4 subfields (18). Work examining hippocampal subregion and subfield volumes in LLD is comparatively sparse with negative studies (19), although some associate subclinical depression with smaller CA1 and subiculum volumes (20). Shape analyses in LLD reported bilateral localized contractions in the left CA4/dentate gyrus subregion (21) and in anterior CA1-CA3 subfields (22). This variability in results may reflect small sample sizes, differences in image segmentation methodology, or inconsistent subregion and subfield definitions.
Relevant to LLD, advancing age is associated with smaller hippocampal volumes and rates of hippocampal atrophy increase from midlife onward (23). Cognitively intact populations exhibit negative relationships between age and volume of the hippocampal head, body and subiculum (24, 25). A large study across the lifespan reported that all subfields exhibited smaller volumes with greater age, except for CA1 and the subiculum which exhibited relatively preserved volumes (26). Volume changes in the subiculum occur early in the course of all-cause dementia and AD (27), while older individuals who were APOE ε4 allele carriers exhibited smaller hippocampal molecular layers and thinning of CA1 (28, 29). We have previously hypothesized that depression is associated with accelerated brain aging (2, 30), and our previous work suggests that age has a more negative effect on medial temporal lobe volumes in individuals with LLD (19).
The purpose of this study was to determine whether LLD was associated with localized differences in specific hippocampal subregions or subfields compared to never-depressed controls. We additionally examined whether the effect of age on cross-sectional hippocampal volume measures differed in individuals with LLD. Investigating hippocampal structure beyond total volume is important, as this may help to differentiate hippocampal microstructural findings in depression from the patterns seen in neurodegenerative conditions, such as AD. A better understanding of microstructural hippocampal alterations in LLD also may suggest targets for neuroplasticity-based treatments. Based on past work, we hypothesized that individuals with LLD would exhibit smaller hippocampal volumes compared to non-depressed peers (7, 8), including smaller volumes of the hippocampus head, CA1, CA3, and dentate gyrus subfields (20–22), and would also exhibit a greater negative effect of age (19) on hippocampal volumes.
As LLD is heterogeneous (2), we additionally tested for differences in clinically relevant depression subgroups. First, we tested for differences between individuals with LLD experiencing a current depressive disorder (current LLD) and those who had remitted from a depressive episode within the last 4 months (remitted LLD). Larger hippocampal volumes are associated with a better response to antidepressant medications (6, 31), while successful antidepressant treatment may also increase hippocampal volumes (32, 33); thus, we hypothesized that hippocampal volumes (including in subregions and subfields) would be increased in remitted compared to current LLD. Second, we tested for differences in individuals with LLD with early vs. late-onset of their first depressive episode (EOD vs. LOD). These subgroups have different etiological factors contributing to depression vulnerability (2). When compared with EOD, individuals with LOD have greater risk for cognitive impairment (34, 35) and past work suggests they exhibit smaller hippocampal and medial temporal volumes (22, 36), which we hypothesized we would see in our sample (including in subregions and subfields).
METHODS
Eligibility Criteria
Participants enrolled in one of six studies with similar entry criteria recruiting individuals either with current or remitted depression or never-depressed controls without a psychiatric history (see Supplemental Materials for individual study details). Common inclusion criteria across studies specified age 60 years or older, English fluency, and absence of dementia on cognitive screening measures (Mini-Mental State Examination (MMSE) (37) ≥22 to 27 depending on study, Montreal Cognitive Assessment (MoCA) (38) ≥24, or Short Blessed Test (SBT)(39) ≤9). Eligible score ranges on these cognitive screens allowed for enrollment of participants with normal performance or mild cognitive symptoms (40). Currently depressed participants met DSM-IV/5 criteria (depending on study; see Supplemental Materials) for current Major Depressive Disorder (MDD) via structured interview (MINI+/SCID) (41, 42) with Montgomery-Asberg Depression Rating Scale (MADRS) (43) score≥15. Remitted depressed participants met DSM-5 criteria for recurrent MDD (partial or full remission), with MADRS≤10 They had to have remitted from a documented major depressive episode in the last 4 months (40). Control participants had no history of psychiatric diagnosis on structured interview with MADRS≤8, plus no history of psychotropic medication use for psychiatric disorder indications.
Common exclusion criteria included: 1) other psychiatric disorders, besides anxiety symptoms occurring during depressive episodes; 2) acute suicidal ideation; 3) current or past psychosis; 4) known primary neurological disorders (e.g., brain tumor, stroke, dementia diagnosis); 5) history of alcohol or drug use disorder within previous 12 months; 6) MRI contraindications; or 7) electroconvulsive therapy in last 6 months.
Participants were recruited via community advertisement and clinician referrals. Each study was approved by the Vanderbilt Institutional Review Board and written informed consent was provided by all participants prior to starting any study procedures.
Clinical Assessment:
At enrollment, participants provided demographic information and completed cognitive screening measures (MMSE, MoCA, or SBT). A study psychiatrist or psychiatric nurse practitioner confirmed diagnoses arising from the structured interview and assessed age of initial onset of depressive episodes via clinical interview. Medical comorbidity was quantified using the Cumulative Illness Rating Scale for Geriatrics (CIRS-G) (44).
Antidepressant use varied across studies. Protocols enrolling currently depressed participants required that any antidepressant medications be withdrawn at least two weeks prior to neuroimaging. However, neuroimaging occurred without a medication withdrawal in the protocol enrolling recently remitted (within the last 4 months) depressed participants (40). Thus, current LLD were not taking antidepressants, while remitted LLD were on antidepressants.
MRI Acquisition and Analyses:
All participants were scanned on research-dedicated 3.0T MRI. Scanner type and pulse sequences varied by study (Supplemental Materials). Analyses used whole-brain T1-weighted MPRAGE images. Following published methods, hippocampal total, subregion, and subfield volumes were extracted by hemisphere using the hippocampal module (45) in FreeSurfer7 (46, 47). This segmentation resulted in total hippocampal volumes, subregion volumes (head, body, tail), and subfield volumes (CA1, CA2/CA3, CA4/dentate gyrus, the molecular layer, the presubiculum and subiculum). Total intracranial volume (ICV) was measured using Spatially Localized Atlas Network Tiles (SLANT) (48, 49).
As neuroimaging data were gathered across different studies and sites (in the case of study 6) using different MRI scanners and pulse sequences, ComBat was used to harmonize data and remove site-related differences while preserving biological variance (50). For the current study, diagnosis, age, and sex were designated as biological variables and the NeuroCombat (version 1.0.13) package in R was applied.
STATISTICAL ANALYSES
Analyses were conducted using R (version 4.2.3). Descriptive statistics were generated using the gtsummary package (version 1.7.2). Continuous variables were compared using Wilcoxon rank sum test (by diagnostic category [LLD vs. control]) and Kruskal-Wallis test (depression status [current LLD vs. remitted LLD vs. control]), while categorical variables were compared using Pearson’s Chi-squared test. Similar to past approaches (19, 21, 22) and given asymmetry in hippocampus structure and function (11), we examined volumes in each hemisphere separately but did not have specific laterality hypotheses.
Primary analyses employed linear regressions to evaluate the effect of diagnostic group (LLD vs. controls) on hippocampal total, subregion, and subfield volumes. All models were adjusted for covariates including age, sex, and ICV. It is recommended that after NeuroCombat harmonization, the same covariates are used in statistical analyses to promote interpretability of results. Analyses of subregion and subfield volumes were adjusted for multiple comparisons using false discovery rate (FDR) correction at α=0.05. Subsequent analyses examined whether the effect of age on hippocampal volumes differed between diagnostic groups. These analyses expanded previous models by adding an interaction term between age and diagnostic group.
We conducted exploratory analyses examining differences within LLD subgroups. First, using the F-test for nested models, we examined effects of depression status on hippocampal volumes across three levels (current LLD vs. recently remitted LLD vs. controls). Statistically significant primary group effects were further explored using post-hoc comparisons between the three groups. Second, linear regressions examined age of onset effects at two levels (EOD vs. LOD). Consistent with our past approaches (51, 52), EOD was defined as an initial depressive episode occurring before age 50 years, and LOD defined as initial onset at age 50 years or older. All models adjusted for age, sex, and ICV. Finally, by expanding these models, we additionally tested for interactive effects between age and diagnostic subgroup: (a) current LLD vs. recently remitted LLD vs. controls; and (b) EOD vs. LOD.
RESULTS
Analyses included 400 older adults, 260 with LLD and 140 never-depressed controls (Table 1). The sample had a mean age of 67.02 (±5.17) years, consisted of 64% women (256/400), and were overall highly educated (mean education=15.81 (±2.39) years). The sample was primarily White (85%; 340/400), with 15% representation from individuals of underrepresented backgrounds (8.5% Black, 6.3% multiracial or of other racial backgrounds) and 3.3% reporting Hispanic ethnicity. There were no significant differences in age, sex, or performance on cognitive screeners between LLD and control groups, but the LLD group reported greater depression severity via the MADRS, less total years of education, greater CIRS-G medical morbidity, and were more likely to be White (Table 1). In the LLD sample, 39.6% were remitted at evaluation and 75.2% of participants reported early-life depression onset. In the 3-group comparisons, controls had more education that both current and remitted LLD groups (who did not differ). Medical morbidity was highest in remitted LLD and depression severity was highest in current LLD. ComBat-adjusted hippocampal volumes across groups are displayed in Table 2.
Table 1.
Diagnostic group demographic characteristics.
| Controls N=140 | LLD N=260 | Current LLD N=157 | Remitted LLD N=103 | 3-group differences | |
|---|---|---|---|---|---|
|
| |||||
| Age (years) | 67.32 (5.46) | 66.86 (5.01) | 66.78 (5.27) | 66.99 (4.63) | ns |
| Sex (% women) | 61.0% (85) | 66.0% (171) | 61.1% (96) | 72.8% (75) | ns |
| Race (% from under-represented backgrounds) | 20.0% (28) | 12.0% (31)** | 10.2% (16) | 14.6 % (15) | ns |
| Ethnicity (% Hispanic) | 2.2% (3) | 3.8% (10) | 3.8% (6) | 3.9% (4) | ns |
| Education (years) | 16.44 (2.01) | 15.48 (2.51)*** | 15.52 (2.19) | 15.42 (2.94) | Controls > LLD |
| Medical morbidity (CIRS-G) | 4.62 (2.69) | 7.34 (3.89)*** | 6.85 (4.08) | 8.11 (3.44) | Remit > Cur > Controls |
| Depression severity (MADRS) | 0.87 (1.30) | 17.95 (10.76)*** | 25.52 (5.33) | 5.57 (3.31) | Cur > Remit > Controls |
| Depression status (% remitted) | -- | 39.6% (103) | -- | -- | -- |
| Age of onset (years) | -- | 33.49 (18.31) | 34.51 (19.79) | 31.37 (15.07) | -- |
| Onset group (% early onset) | -- | 75.2% (176) | 71.7% (109) | 81.7% (67) | -- |
| MMSE total (n=214) | 29.18 (1.04) | 29.20 (1.05) | 29.20 (1.05) | -- | -- |
| MoCA total (n=103) | 26.27 (1.40) | 26.17 (2.00) | -- | 26.17 (2.00) | -- |
| SBT total (n=31) | 1.54 (1.66) | 1.33 (1.68) | 1.33 (1.68) | -- | -- |
The late-life depression (LLD) sample includes all participants with LLD, combining current and remitted depression. Age of onset and onset group included 237 and 234 LLD participants total, respectively, due to missing data. Cognitive screeners differed between studies. Analysis of categorical variables utilized Pearson’s chi-square tests, while analyses of continuous variables used Wilcoxen rank sum or Kruskal-Wallis tests. CIRS-G = Cumulative Illness Rating Score for Geriatrics; MADRS = Montgomery-Asberg Depression Rating Scale; MMSE = Mini Mental State Examination; MoCA = Montreal Cognitive Assesment; SBT = Short Blessed Test.
p < 0.05
p < 0.001
ns = not significant.
Table 2:
Hippocampal volumes across study groups
| Diagnostic Comparison | Depression Comparison | Age of Onset Comparison | |||||
|---|---|---|---|---|---|---|---|
|
| |||||||
| Region | LLD (N=260) | Controls (N=140) | Remitted LLD (N=103) | Current LLD (N=157) | Early Onset (N=176) | Late Onset (N=58) | |
|
| |||||||
| LEFT HEMISPHERE | |||||||
|
| |||||||
| Whole | Total | 3.26 (0.39) | 3.33 (0.38) | 3.23 (0.36) | 3.28 (0.41) | 3.28 (0.38) | 3.21 (0.44) |
| Subregions | Head | 1.60 (0.22) | 1.64 (0.20) | 1.59 (0.20) | 1.61 (0.23) | 1.61 (0.21) | 1.58 (0.23) |
| Body | 1.13 (0.14) | 1.15 (0.13) | 1.12 (0.12) | 1.14 (0.15) | 1.13 (0.13) | 1.11 (0.16) | |
| Tail | 0.53 (0.08) | 0.55 (0.08) | 0.53 (0.08) | 0.53 (0.07) | 0.53 (0.07) | 0.53 (0.09) | |
| Subfields | CA1 | 0.60 (0.09) | 0.61 (0.09) | 0.60 (0.08) | 0.60 (0.09) | 0.60 (0.08) | 0.59 (0.09) |
| CA2/3 | 0.20 (0.03) | 0.20 (0.03) | 0.20 (0.03) | 0.20 (0.03) | 0.20 (0.03) | 0.20 (0.03) | |
| CA4/Dentate | 0.50 (0.07) | 0.52 (0.07) | 0.50 (0.06) | 0.51 (0.07) | 0.50 (0.06) | 0.50 (0.07) | |
| Molecular Layer | 0.52 (0.07) | 0.53 (0.06) | 0.52 (0.06) | 0.53 (0.07) | 0.53 (0.07) | 0.52 (0.07) | |
| Presubiculum | 0.30 (0.05) | 0.31 (0.04) | 0.29 (0.04) | 0.30 (0.05) | 0.30 (0.05) | 0.29 (0.05) | |
| Subiculum | 0.42 (0.06) | 0.42 (0.05) | 0.41 (0.06) | 0.42 (0.06) | 0.42 (0.06) | 0.41 (0.06) | |
|
| |||||||
| RIGHT HEMISPHERE | |||||||
|
| |||||||
| Whole | Total | 3.36 (0.40) | 3.45 (0.36) | 3.35 (0.38) | 3.36 (0.41) | 3.35 (0.39) | 3.34 (0.42) |
| Subregions | Head | 1.66 (0.22) | 1.70 (0.20) | 1.66 (0.21) | 1.65 (0.23) | 1.66 (0.22) | 1.66 (0.23) |
| Body | 1.15 (0.14) | 1.17 (0.13) | 1.15 (0.13) | 1.14 (0.14) | 1.14 (0.14) | 1.13 (0.15) | |
| Tail | 0.55 (0.08) | 0.57 (0.08) | 0.55 (0.08) | 0.56 (0.08) | 0.55 (0.07) | 0.55 (0.08) | |
| Subfields | CA1 | 0.63 (0.09) | 0.65 (0.09) | 0.64 (0.08) | 0.63 (0.09) | 0.63 (0.09) | 0.63 (0.09) |
| CA2/3 | 0.22 (0.03) | 0.23 (0.03) | 0.22 (0.03) | 0.22 (0.03) | 0.22 (0.03) | 0.22 (0.03) | |
| CA4/Dentate | 0.53 (0.07) | 0.54 (0.07) | 0.54 (0.07) | 0.53 (0.07) | 0.53 (0.07) | 0.54 (0.07) | |
| Molecular Layer | 0.54 (0.07) | 0.55 (0.06) | 0.54 (0.06) | 0.54 (0.07) | 0.54 (0.07) | 0.54 (0.07) | |
| Presubiculum | 0.28 (0.04) | 0.28 (0.03) | 0.28 (0.04) | 0.28 (0.04) | 0.28 (0.04) | 0.27 (0.04) | |
| Subiculum | 0.42 (0.06) | 0.42 (0.05) | 0.41 (0.05) | 0.42 (0.06) | 0.42 (0.05) | 0.41 (0.06) | |
Notes. Values reported are after ComBat (but not covariate) adjustment. LLD = late-life depression; CA = cornus ammonis.
Diagnostic differences: Volumetric differences in LLD vs. control participants
Even after regressing out the effects of age (which was a significant covariate in all of the models, p’s<0.001) and adjusting for multiple comparisons, compared to the control group, participants with LLD exhibited significantly smaller total hippocampal volumes bilaterally (left: estimate=−70.21 (SE=30.23), t=−2.32, p=0.021; right: estimate=−94.34 (SE=29.40), t=−3.21, p=0.001; Table 3). These findings were driven by group differences in the bilateral hippocampal head and the right hippocampal body and tail volumes, with LLD exhibiting significantly smaller volumes than controls. For subfield volumes, there were statistically significant group differences in the right CA1 and right molecular layer regions, with LLD exhibiting smaller volumes than controls (Table 3).
Table 3.
Comparisons in volumes between individuals with and without late-life depression (LLD)
| Region | Estimate (SE) | t-value | Unadjusted p value | Adjusted p value | |
|---|---|---|---|---|---|
|
| |||||
| LEFT HEMISPHERE | |||||
|
| |||||
| Whole | Total | −70.21 (30.23) | −2.32 | 0.021 | 0.021 |
| Subregions | Head | −39.60 (16.69) | −2.37 | 0.018 | 0.027 |
| Body | −17.53 (11.00) | −1.59 | 0.112 | 0.112 | |
| Tail | −13.42 (7.27) | −1.85 | 0.066 | 0.079 | |
| Subfields | CA1 | −14.95 (7.23) | −2.07 | 0.039 | 0.067 |
| CA2/3 | −3.08 (2.69) | −1.14 | 0.253 | 0.276 | |
| CA4 / Dentate | −11.69 (5.33) | −2.19 | 0.029 | 0.067 | |
| Molecular Layer | −11.14 (5.32) | −2.10 | 0.037 | 0.067 | |
| Presubiculum | −6.22 (3.97) | −1.57 | 0.118 | 0.157 | |
| Subiculum | −2.61 (4.65) | −0.56 | 0.576 | 0.576 | |
|
| |||||
| RIGHT HEMISPHERE | |||||
|
| |||||
| Whole | Total | −94.34 (29.40) | −3.21 | 0.001 | 0.003 |
| Subregions | Head | −48.66 (16.63) | −2.93 | 0.004 | 0.022 |
| Body | −27.44 (10.88) | −2.52 | 0.012 | 0.024 | |
| Tail | −18.55 (7.09) | −2.62 | 0.009 | 0.024 | |
| Subfields | CA1 | −21.86 (7.30) | −3.00 | 0.003 | 0.020 |
| CA2/3 | −5.96 (2.72) | −2.19 | 0.029 | 0.067 | |
| CA4 / Dentate | −12.22 (5.24) | −2.33 | 0.020 | 0.067 | |
| Molecular Layer | −15.14 (5.14) | −2.94 | 0.003 | 0.020 | |
| Presubiculum | −4.72 (3.36) | −1.41 | 0.161 | 0.193 | |
| Subiculum | −8.56 (4.60) | −1.86 | 0.064 | 0.095 | |
Notes. Models adjusted for age, sex, and total intracranial volume. Adjusted p value reflects FDR-correction. LLD = late-life depression; CA = cornus ammonis. Bold indicates significance at adjusted p < 0.05. Comparison group is controls.
Interactive effect of age by diagnostic group (LLD vs. control) on hippocampal volumes
After multiple comparison adjustment, there were no significant age by diagnostic group interactions for left (estimate=−10.25 (SE=5.71), t=−1.80, p=0.123) or right (estimate=−8.60 (SE=5.56), t=−1.55, p=0.123) hemisphere total hippocampal volumes. Age by diagnostic group interactions were also not significant for any subregion or subfield volume (p‘s>0.190; Supplemental Table 1).
Subgroup analyses: Comparison of current LLD vs. remitted LLD vs. controls
Group differences:
In the LLD sample, 157 participants had current LLD and 103 had remitted LLD. After controlling for covariates (including age, which was significant in all of the models, p’s<0.001) and multiple comparison correction, models testing for volume differences between the three groups (current LLD vs. remitted LLD vs. controls) revealed a significant primary group effect for total hippocampus volume in the right (χ2 (2, N=400)=12.03, p=0.005), but not left (χ2 (2, N=400)=5.40, p=0.067) hemisphere. In post-hoc analyses (Table 4), current LLD exhibited smaller right total hippocampal volumes than the control group, but not the remitted LLD group. Total hippocampal volumes did not significantly differ between remitted LLD and controls.
Table 4.
Volumetric differences in hippocampus whole, subregions, and subfields between currently depressed, remitted depressed, and controls
| Omnibus (Primary effect of group) | Remitted LLD vs Controls | Current LLD vs Controls | Remitted LLD vs Current LLD | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
| ||||||||||
| Region | X2 value | Unadjusted p value | Adjusted p value | Estimate (SE) | t value / p value | Estimate (SE) | t value / p value | Estimate (SE) | t value / p value | |
|
| ||||||||||
| LEFT HEMISPHERE | ||||||||||
|
| ||||||||||
| Whole | Total | 5.40 | 0.067 | 0.067 | −66.83 (37.76) | −1.77 / 0.078 | −72.86 (33.68) | −2.16 / 0.031 | 6.04 (36.88) | 0.16 / 0.870 |
| Subregions | Head | 5.83 | 0.054 | 0.081 | −33.91 (20.85) | −1.63 / 0.105 | −43.51 (18.59) | −2.34 / 0.020 | 9.60 (20.36) | 0.47 / 0.638 |
| Body | 2.58 | 0.275 | 0.275 | −18.72 (13.73) | −1.36 / 0.173 | −17.05 (12.25) | −1.39 / 0.165 | −1.67 (13.41) | −0.13 / 0.901 | |
| Tail | 3.53 | 0.171 | 0.206 | −10.48 (9.10) | −1.15 / 0.250 | −15.02 (8.11) | −1.85 / 0.065 | 4.55 (8.89) | 0.51 / 0.609 | |
| Subfields | CA1 | 4.93 | 0.085 | 0.185 | −10.51 (9.03) | −1.16 / 0.245 | −17.84 (8.05) | −2.22 / 0.027 | 7.33 (8.82) | 0.83 / 0.406 |
| CA2/3 | 3.02 | 0.221 | 0.266 | −0.44 (3.36) | −0.13 / 0.896 | −4.76 (3.00) | −1.59 / 0.113 | 4.32 (3.28) | 1.32 / 0.189 | |
| CA4 / Dentate | 4.76 | 0.092 | 0.185 | −10.99 (6.65) | −1.65 / 0.100 | −12.09 (5.93) | −2.04 / 0.042 | 1.10 (6.50) | 0.17 / 0.865 | |
| Molecular Layer | 4.41 | 0.110 | 0.189 | −10.44 (6.64) | −1.57 / 0.117 | −11.66 (5.92) | −1.97 / 0.050 | 1.22 (6.49) | 0.19 / 0.851 | |
| Presubiculum | 3.82 | 0.148 | 0.197 | −9.67 (4.95) | −1.96 / 0.051 | −4.10 (4.41) | −0.93 / 0.353 | −5.57 (4.83) | −1.15 / 0.250 | |
| Subiculum | 0.39 | 0.821 | 0.821 | −3.56 (5.81) | −0.61 / 0.540 | −2.11 (5.18) | −0.41 / 0.684 | −1.45 (5.68) | −0.26 / 0.799 | |
|
| ||||||||||
| RIGHT HEMISPHERE | ||||||||||
|
| ||||||||||
| Whole | Total | 12.03 | 0.002 | 0.005 | −63.79 (36.67) | −1.74 / 0.083 | −113.40 (32.70) | −3.47 / <0.001 | 49.58 (35.82) | 1.38 / 0.167 |
| Subregions | Head | 12.05 | 0.002 | 0.014 | −24.71 (20.73) | −1.19 / 0.234 | −63.55 (18.49) | −3.44 / <0.001 | −38.84 (20.24) | 1.92 / 0.056 |
| Body | 8.85 | 0.012 | 0.036 | −14.53 (13.57) | −1.07 / 0.285 | −35.72 (12.10) | −2.95 / 0.003 | 21.19 (13.25) | 1.60 / 0.111 | |
| Tail | 6.76 | 0.034 | 0.068 | −20.15 (8.87) | −2.27 / 0.024 | −17.28 (7.91) | −2.19 / 0.029 | −2.87 (8.66) | −0.33 / 0.741 | |
| Subfields | CA1 | 14.45 | <0.001 | 0.008 | −8.87 (9.08) | −0.98 / 0.329 | −29.95 (8.10) | −3.70 / <0.001 | 21.08 (8.87) | 2.38 / 0.018 |
| CA2/3 | 12.70 | 0.002 | 0.008 | −0.27 (3.37) | −0.08 / 0.936 | −9.56 (3.01) | −3.18 / 0.002 | 9.29 (3.29) | 2.82 / 0.005 | |
| CA4 / Dentate | 11.85 | 0.003 | 0.008 | −2.02 (6.53) | −0.31 / 0.757 | −18.44 (5.82) | −3.17 / 0.002 | 16.42 (6.37) | 2.58 / 0.010 | |
| Molecular Layer | 11.86 | 0.003 | 0.008 | −8.07 (6.41) | −1.26 / 0.208 | −19.54 (5.71) | −3.42 / <0.001 | 11.47 (6.26) | 1.83 / 0.068 | |
| Presubiculum | 2.48 | 0.289 | 0.316 | −6.43 (4.19) | −1.53 / 0.126 | −3.89 (3.74) | −1.04 / 0.299 | −2.54 (4.10) | −0.62 / 0.535 | |
| Subiculum | 4.05 | 0.132 | 0.197 | −6.09 (5.74) | −1.06 / 0.289 | −10.27 (5.12) | −2.01 / 0.045 | 4.18 (5.60) | 0.75 / 0.456 | |
Notes. Models adjusted for age, sex, and total intracranial volume. Adjusted p value reflects FDR-correction. LLD = late-life depression; CA = cornus ammonis. Bold indicates significance of adjusted p < 0.05 for omnibus model. Comparison group is either controls or current LLD.
The right hemisphere findings in current LLD were driven by findings in right hippocampal head (χ2 (2, N=400)=12.06, p=0.014) and body (χ2 (2, N=400)=8.85, p=0.035) volumes, with the only significant post-hoc group differences being that current LLD exhibited smaller volumes than controls (Table 4). No other group differences were evident for other subregions (Table 4). When examining subfield volumes, significant group differences were seen in right hemisphere CA1, CA3, CA4/dentate gyrus, and molecular layer volumes (Table 4). In post hoc analyses, volumes were smaller in current LLD compared to controls (p’s≤0.002). For right CA1, CA3, and CA4/dentate gyrus, current LLD volumes were also significantly smaller than remitted LLD (p’s<0.018). Subfield volumes in remitted LLD did not differ from controls in any analysis (p’s>0.208; Table 4).
Interactive effect of age by diagnostic group:
After multiple comparison correction, there were no significant age by group interactions for left (χ2 (2, N=400)=6.15, p=0.093) or right (χ2 (2, N=400)=4.55, p=0.103) hemisphere total hippocampal volumes. Age by group interaction terms were also not significant for any subregion or subfield volume (p‘s>0.102; Supplemental Table 2).
Subgroup analyses: Comparison of early vs. late-onset LLD
The LLD sample included 176 individuals with EOD and 58 individuals with LOD. After controlling for covariates (including age, which was significant in all of the models, p’s<0.001), compared to EOD, the LOD group did not statistically differ in total hippocampal volume for the left (estimate=−44.13 (SE=39.451), t=−1.12, p=0.265) or right (estimate=17.80 (SE=39.45), t=0.45, p=0.652) hemisphere. Similarly, there were no statistically significant group differences in hippocampal subregion and subfield volumes (data not shown). Finally, we did not observe any statistically significant interactive effects between age and onset group (EOD or LOD) on any hippocampal volume measures (data not shown). We conducted additional secondary analyses investigating age of depression onset effects on hippocampal volumes based on the distribution in our sample: onset≤age 24 (n=94), onset age 25 to 54 (n=93), and onset≥55 (n=43). After covariate and multiple comparison adjustments, there were no significant differences between these age of onset groups on any hippocampal volume measure (p’s≥0.796).
Exploratory analyses: Other factors affecting brain health.
As level of education and medical burden may influence hippocampal volumes, we re-ran the 2- and 3-group depression status analyses adding both as covariates. This involved re-running NeuroCombat harmonization first with the same covariates being used in the statistical analyses. After adjusting for covariates and multiple comparisons, LLD had significantly smaller total right hippocampal volumes (estimate=−86.50 (SE=32.15), t=−2.60, p=0.015) than controls (see Supplemental Table 3). Left hemisphere and right subregion and subfield hippocampal volumes were not significant after adjustment for multiple comparisons.
In models testing for volume differences between the 3-groups (current LLD vs. remitted LLD vs. controls; Supplemental Table 4), the significant primary group effect for total hippocampal volume in the right hemisphere remained (χ2 (2, N=394)=9.04, p=0.022), with the current LLD group exhibiting smaller total hippocampal volumes than the control group. Subregion volumes were not significant after adjustment for multiple comparisons. The pattern of findings across the right subfields (CA1, CA3, CA4/dentate gyrus, and molecular layer) was similar to the analyses that did not control for education and medical burden.
DISCUSSION
This study is amongst the larger cross-sectional examinations of hippocampal volumes in LLD. It expands work in previous meta-analyses (7, 8) by examining volumes of hippocampal subregions and subfields. The observed differences between LLD and controls in total hippocampal volumes are concordant with broader literature associating depression with smaller hippocampal volumes in both adult (3, 4) and geriatric populations (7, 8). When separating currently depressed from recently remitted groups, we observed significantly smaller right total hippocampus volume in current LLD compared to controls, with remitted LLD exhibiting mean volumes in-between the currently depressed and control groups. The remitted group had successfully obtained remission within the previous 4 months through algorithm-guided treatment (40), and previous research has shown that smaller hippocampal volumes decrease the likelihood of remission with treatment (6, 31). Thus, by responding to treatment, this subgroup may be enriched towards having larger hippocampal volumes. Alternatively, successful antidepressant treatment and depression remission may increase hippocampal volumes (32, 33), potentially via neurotrophic benefits of antidepressant treatment (53). Given the nature of our sample, it is difficult to disentangle whether effects in the recently remitted group are due to them being on antidepressants and/or the recent absence of symptoms.
We extended past work by examining hippocampal anterior-posterior axis subregions. The anatomical connections and functional properties of the hippocampus differ along this axis, with the anterior head processing emotional, social, and semantic stimuli and the posterior tail processing spatial and visual information (13–15). This is mirrored by hippocampal connectivity with intrinsic functional networks, where the anterior aspect is often included in the Default Mode Network (DMN), while the posterior aspect is part of the visual network. Memory functions are similarly distributed along the anterior-posterior axis, with encoding and associative processes occurring in anterior regions and retrieval in posterior regions (13, 54). In the present study, individuals with LLD exhibited smaller subregion volumes across the axis (head, body, and tail) in the right hemisphere only. When examining current and recently remitted depressed participants separately, right hemisphere head and body in current LLD were significantly smaller than controls but not remitted LLD (who did not significantly differ from controls). For subfields, the right CA1 was smaller in current LLD compared to controls, while right CA1, CA4/dentate gyrus, and molecular layer were smaller in current LLD compared to both controls and remitted LLD (who did not differ from one another). The total and subfield findings remained even after adding additional covariates (education, medical burden) to the models. While some research supports smaller right hippocampal total and subregion volumes in depression (16, 55–57), a more recent systematic review and meta-analysis did not find evidence for hippocampal lateralization in LLD (7). This lack of consensus may be partly due to the heterogeneous nature of LLD in terms of etiology, underlying mechanisms, and clinical characteristics. Methodological differences across studies, including MRI acquisition and defining hippocampal boundaries, is also contributory. While some conditions, such as AD, tend to affect left hemisphere hippocampal volumes to a greater extent (58–60), there is inconsistency in hippocampal lateralization in LLD and caution is advised when interpreting any hemispheric differences.
These regional findings may have clinical significance, as LLD is characterized by a range of emotional, social, and cognitive difficulties (1, 2). Reduced hippocampal tail volumes may be particularly problematic, as they are associated with poorer spatial processing and are predictive of a poorer antidepressant response in adult MDD (31). Reduced hippocampal head volumes may adversely affect memory encoding and processing of emotional, motivational, or social information (14). Reduced head and tail volumes may both be related to cognitive outcomes in LLD, as both anterior and posterior hippocampal functional networks are broadly affected by AD (15, 61). We were unable to test these brain-behavior hypotheses in the current dataset, as contributing studies included different cognitive and behavioral measurements.
Contrary to some past work (22, 62), we did not observe differences in hippocampal total, subregion, or subfield volumes based on age of initial depression onset. This supports the hypothesis that depression across the lifespan (regardless of EOD or LOD status) may have deleterious effects, with recurrent depressive episodes increasing vulnerability to poor longitudinal outcomes (2, 63), including negative effects on hippocampal structure. It is important to note that the individual studies that comprise the current study were not powered to detect differences between EOD and LOD, as seen by the imbalance is subsample sizes. Thus, while the current study had a negative finding for age of initial depression onset, this may not be replicated in other research.
In contrast to our hypotheses, we did not find significant age by depression subgroup interactions for any of the hippocampal volumetric measurements. The main effect of age on hippocampal volumes was significant for all models; however, the effect of age did not differ between LLD and control groups or for any of the depression subgroups. While other research suggests that LLD may be associated with accelerated brain aging (30, 64), it may be that other brain regions age faster in LLD (such as the frontal lobe) and the hippocampus may be less vulnerable to accelerated aging in the absence of neuropathology. Cognitive impairment in LLD is not generally thought to be primarily related to AD biomarkers (65, 66), the latter of which would be more susceptible to hippocampal atrophy. Given LLD’s heterogeneity, there may be LLD subgroups that demonstrate evidence of accelerated hippocampal aging (e.g., those with amnestic memory profiles), but this is not evident when examining the group as whole.
Despite strengths including the large sample size and adjustment for multiple comparisons, this study has additional limitations beyond its cross-sectional nature. Notably, there was heterogeneity in antidepressant use across participants in the contributing studies. While the recently remitted cohort were all taking antidepressant medications, there was heterogeneity of antidepressant regimens (40). This group also had a longer amount of time to experience the potential “benefits” of antidepressants on hippocampal volume, which may have influenced results beyond the absence of clinically significantly depressive symptoms. Likewise, while currently depressed individuals underwent a two-week washout of ineffective antidepressant medications prior to neuroimaging, it is possible that residual antidepressant effects may have persisted and influenced our results. Our cohort also had minimal representation at age ranges above 80 years and a smaller proportion of participants with LOD, so we may have been under-powered to detect effects. It is worth acknowledging that other mental health symptoms beyond depression are associated with subfield differences, including anxiety and worry (67), but such domains were not consistently quantified across studies used in the current analysis. We additionally did not have consistent data across contributing samples about persistence or chronicity of the current depressive episode, which may be associated with volumetric differences (68). We attempted to minimize significant cognitive impairment in our studies via the cognitive screening cut-offs; however, we cannot rule out whether some individuals were in the early phases of a neurodegenerative process, which could affect findings. Finally, there is evidence to suggest extraction of some hippocampal subfield volumes at 1 mm are less reliable (specifically, CA4/dentate gyrus and molecular layer) and need further validation with higher resolution (45), which makes for cautious interpretation of these specific findings.
In conclusion, in this large multi-study sample, individuals with LLD exhibited smaller total, subregion, and subfield hippocampal volumes, particularly in the right hemisphere. This was most apparent in current LLD compared to controls. The lack of differences between remitted LLD and controls is likely influenced by sample-specific confounding factors. We did not find significant age by diagnostic group interactions on hippocampal volumes across any of the analyses. This cross-sectional work supports diagnostic, but not accelerated aging, effects of depression on the hippocampus. Longitudinal studies are needed to examine trajectory of changes and the effects of depression remission or relapse status on such trajectories and long-term outcomes (69), including both depression recurrence and cognitive decline (40).
Supplementary Material
ACKNOWLEDGEMENTS:
This work was supported by the Vanderbilt Faculty Research Scholars Program and National of Institute of Health grants R01 MH102246, R21 MH099218, R01 MH123662, R61 MH122464, R01 MH121619, R01 MH121620, and R01 MH121384, and the National Center for Advancing Translational Sciences grants UL1 TR000445 andUL1 TR002243.
Footnotes
DECLARATION OF INTERESTS: Dr. Ajilore is the co-founder of KeyWise AI. He serves as a consultant for Otsuka and has served as a consultant for Sage Therapeutics. He is on the advisory board of Embodied Labs and Blueprint. All other authors deny any conflict of interest with this work.
REFERENCES
- 1.Taylor WD (2014): Clinical practice. Depression in the elderly. N Engl J Med. 371:1228–1236. [DOI] [PubMed] [Google Scholar]
- 2.Szymkowicz SM, Gerlach AR, Homiack D, Taylor WD (2023): Biological factors influencing depression in later life: role of aging processes and treatment implications. Transl Psychiatry. 13:160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schmaal L, Veltman DJ, van Erp TG, Samann PG, Frodl T, Jahanshad N, et al. (2016): Subcortical brain alterations in major depressive disorder: findings from the ENIGMA Major Depressive Disorder working group. Mol Psychiatry. 21:806–812. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Arnone D, McIntosh AM, Ebmeier KP, Munafo MR, Anderson IM (2012): Magnetic resonance imaging studies in unipolar depression: systematic review and meta-regression analyses. Eur Neuropsychopharmacol. 22:1–16. [DOI] [PubMed] [Google Scholar]
- 5.Szymkowicz SM, Woods AJ, Dotson VM, Porges EC, Nissim NR, O'Shea A, et al. (2019): Associations between subclinical depressive symptoms and reduced brain volume in middle-aged to older adults. Aging & mental health. 23:819–830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gerlach AR, Karim HT, Pecina M, Ajilore O, Taylor WD, Butters MA, et al. (2022): MRI predictors of pharmacotherapy response in major depressive disorder. Neuroimage Clin. 36:103157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Geerlings MI, Gerritsen L (2017): Late-Life Depression, Hippocampal Volumes, and Hypothalamic-Pituitary-Adrenal Axis Regulation: A Systematic Review and Meta-analysis. Biol Psychiatry. 82:339–350. [DOI] [PubMed] [Google Scholar]
- 8.Sexton CE, Mackay CE, Ebmeier KP (2013): A systematic review and meta-analysis of magnetic resonance imaging studies in late-life depression. Am J Geriatr Psychiatry. 21:184–195. [DOI] [PubMed] [Google Scholar]
- 9.Steffens DC, Payne ME, Greenberg DL, Byrum CE, Welsh-Bohmer KA, Wagner HR, et al. (2002): Hippocampal volume and incident dementia in geriatric depression. Am J Geriatr Psychiatry. 10:62–71. [PubMed] [Google Scholar]
- 10.De Winter FL, Emsell L, Bouckaert F, Claes L, Jain S, Farrar G, et al. (2017): No Association of Lower Hippocampal Volume With Alzheimer's Disease Pathology in Late-Life Depression. Am J Psychiatry. 174:237–245. [DOI] [PubMed] [Google Scholar]
- 11.Robinson JL, Barron DS, Kirby LA, Bottenhorn KL, Hill AC, Murphy JE, et al. (2015): Neurofunctional topography of the human hippocampus. Hum Brain Mapp. 36:5018–5037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Small SA, Schobel SA, Buxton RB, Witter MP, Barnes CA (2011): A pathophysiological framework of hippocampal dysfunction in ageing and disease. Nat Rev Neurosci. 12:585–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Beason-Held LL, Shafer AT, Goh JO, Landman BA, Davatzikos C, Viscomi B, et al. (2021): Hippocampal activation and connectivity in the aging brain. Brain Imaging Behav. 15:711–726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Grady CL (2020): Meta-analytic and functional connectivity evidence from functional magnetic resonance imaging for an anterior to posterior gradient of function along the hippocampal axis. Hippocampus. 30:456–471. [DOI] [PubMed] [Google Scholar]
- 15.Dautricourt S, de Flores R, Landeau B, Poisnel G, Vanhoutte M, Delcroix N, et al. (2021): Longitudinal Changes in Hippocampal Network Connectivity in Alzheimer's Disease. Ann Neurol. 90:391–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Malykhin NV, Carter R, Seres P, Coupland NJ (2010): Structural changes in the hippocampus in major depressive disorder: contributions of disease and treatment. J Psychiatry Neurosci. 35:337–343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chu Z, Yuan L, Lian K, He M, Lu Y, Cheng Y, et al. (2024): Reduced gray matter volume of the hippocampal tail in melancholic depression: evidence from an MRI study. BMC Psychiatry. 24:183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sun Y, Hu N, Wang M, Lu L, Luo C, Tang B, et al. (2023): Hippocampal subfield alterations in schizophrenia and major depressive disorder: a systematic review and network meta-analysis of anatomic MRI studies. J Psychiatry Neurosci. 48:E34–E49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Taylor WD, Deng Y, Boyd BD, Donahue MJ, Albert K, McHugo M, et al. (2020): Medial temporal lobe volumes in late-life depression: effects of age and vascular risk factors. Brain Imaging Behav. 14:19–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lee J, Ju G, Park H, Chung S, Son JW, Shin CJ, et al. (2022): Hippocampal Subfields and White Matter Connectivity in Patients with Subclinical Geriatric Depression. Brain Sci. 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhao Z, Taylor WD, Styner M, Steffens DC, Krishnan KR, MacFall JR (2008): Hippocampus shape analysis and late-life depression. PLoS One. 3:e1837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ballmaier M, Narr KL, Toga AW, Elderkin-Thompson V, Thompson PM, Hamilton L, et al. (2008): Hippocampal morphology and distinguishing late-onset from early-onset elderly depression. Am J Psychiatry. 165:229–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fraser MA, Shaw ME, Cherbuin N (2015): A systematic review and meta-analysis of longitudinal hippocampal atrophy in healthy human ageing. Neuroimage. 112:364–374. [DOI] [PubMed] [Google Scholar]
- 24.Malykhin NV, Huang Y, Hrybouski S, Olsen F (2017): Differential vulnerability of hippocampal subfields and anteroposterior hippocampal subregions in healthy cognitive aging. Neurobiol Aging. 59:121–134. [DOI] [PubMed] [Google Scholar]
- 25.Wolf D, Fischer FU, de Flores R, Chetelat G, Fellgiebel A (2015): Differential associations of age with volume and microstructure of hippocampal subfields in healthy older adults. Hum Brain Mapp. 36:3819–3831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bussy A, Plitman E, Patel R, Tullo S, Salaciak A, Bedford SA, et al. (2021): Hippocampal subfield volumes across the healthy lifespan and the effects of MR sequence on estimates. Neuroimage. 233:117931. [DOI] [PubMed] [Google Scholar]
- 27.Evans TE, Adams HHH, Licher S, Wolters FJ, van der Lugt A, Ikram MK, et al. (2018): Subregional volumes of the hippocampus in relation to cognitive function and risk of dementia. Neuroimage. 178:129–135. [DOI] [PubMed] [Google Scholar]
- 28.Dounavi ME, Mak E, Wells K, Ritchie K, Ritchie CW, Su L, et al. (2020): Volumetric alterations in the hippocampal subfields of subjects at increased risk of dementia. Neurobiol Aging. 91:36–44. [DOI] [PubMed] [Google Scholar]
- 29.Kerchner GA, Berdnik D, Shen JC, Bernstein JD, Fenesy MC, Deutsch GK, et al. (2014): APOE epsilon4 worsens hippocampal CA1 apical neuropil atrophy and episodic memory. Neurology. 82:691–697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Christman S, Bermudez C, Hao L, Landman BA, Boyd B, Albert K, et al. (2020): Accelerated brain aging predicts impaired cognitive performance and greater disability in geriatric but not midlife adult depression. Transl Psychiatry. 10:317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Nogovitsyn N, Muller M, Souza R, Hassel S, Arnott SR, Davis AD, et al. (2020): Hippocampal tail volume as a predictive biomarker of antidepressant treatment outcomes in patients with major depressive disorder: a CAN-BIND report. Neuropsychopharmacology. 45:283–291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Arnone D, McKie S, Elliott R, Juhasz G, Thomas EJ, Downey D, et al. (2013): State-dependent changes in hippocampal grey matter in depression. Mol Psychiatry. 18:1265–1272. [DOI] [PubMed] [Google Scholar]
- 33.Enneking V, Leehr EJ, Dannlowski U, Redlich R (2020): Brain structural effects of treatments for depression and biomarkers of response: a systematic review of neuroimaging studies. Psychol Med. 50:187–209. [DOI] [PubMed] [Google Scholar]
- 34.Koenig AM, Bhalla RK, Butters MA (2014): Cognitive functioning and late-life depression. J Int Neuropsychol Soc. 20:461–467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Dillon C, Allegri RF, Serrano CM, Iturry M, Salgado P, Glaser FB, et al. (2009): Late- versus early-onset geriatric depression in a memory research center. Neuropsychiatr Dis Treat. 5:517–526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Gerritsen L, Comijs HC, van der Graaf Y, Knoops AJ, Penninx BW, Geerlings MI (2011): Depression, hypothalamic pituitary adrenal axis, and hippocampal and entorhinal cortex volumes--the SMART Medea study. Biol Psychiatry. 70:373–380. [DOI] [PubMed] [Google Scholar]
- 37.Folstein MF, Folstein SE, McHugh PR (1975): “Mini-mental state” a practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research. 12:189–198. [DOI] [PubMed] [Google Scholar]
- 38.Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, Whitehead V, Collin I, et al. (2005): The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 53:695–699. [DOI] [PubMed] [Google Scholar]
- 39.Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H (1983): Validation of a short Orientation-Memory-Concentration Test of cognitive impairment. Am J Psychiatry. 140:734–739. [DOI] [PubMed] [Google Scholar]
- 40.Taylor WD, Ajilore O, Karim HT, Butters MA, Krafty R, Boyd BD, et al. (2024): Assessing depression recurrence, cognitive burden, and neurobiological homeostasis in late life: Design and rationale of the REMBRANDT Study. J Mood Anxiety Disord. 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. (1998): The Mini-International Neuropsychiatric Inventory (M.I.N.I.): the development and validation of a structured diagnostic interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry. Suppl 20:22–33. [PubMed] [Google Scholar]
- 42.First MB, Williams JBW, Karg RS, Spitzer RL (2015): Structured Clinical Interview for DSM-5 - Research Version. Washington, DC: American Psychiatric Association Publishing. [Google Scholar]
- 43.Montgomery SA, Asberg M (1979): A new depression scale designed to be sensitive to change. British Journal of Psychiatry. 134:382–389. [DOI] [PubMed] [Google Scholar]
- 44.Miller MD, Paradis CF, Houck PR, Mazumdar S, Stack JA, Rifai AH, et al. (1992): Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res. 41:237–248. [DOI] [PubMed] [Google Scholar]
- 45.Iglesias JE, Augustinack JC, Nguyen K, Player CM, Player A, Wright M, et al. (2015): A computational atlas of the hippocampal formation using ex vivo, ultra-high resolution MRI: Application to adaptive segmentation of in vivo MRI. Neuroimage. 115:117–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, et al. (2002): Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron. 33:341–355. [DOI] [PubMed] [Google Scholar]
- 47.Fischl B, Salat DH, van der Kouwe AJ, Makris N, Segonne F, Quinn BT, et al. (2004): Sequence-independent segmentation of magnetic resonance images. Neuroimage. 23 Suppl 1:S69–84. [DOI] [PubMed] [Google Scholar]
- 48.Huo Y, Asman AJ, Plassard AJ, Landman BA (2017): Simultaneous total intracranial volume and posterior fossa volume estimation using multi-atlas label fusion. Hum Brain Mapp. 38:599–616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Huo Y, Xu Z, Xiong Y, Aboud K, Parvathaneni P, Bao S, et al. (2019): 3D whole brain segmentation using spatially localized atlas network tiles. Neuroimage. 194:105–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Fortin JP, Cullen N, Sheline YI, Taylor WD, Aselcioglu I, Cook PA, et al. (2018): Harmonization of cortical thickness measurements across scanners and sites. Neuroimage. 167:104–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Burke J, McQuoid DR, Payne ME, Steffens DC, Krishnan RR, Taylor WD (2011): Amygdala volume in late-life depression: relationship with age of onset. Am J Geriatr Psychiatry. 19:771–776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Riddle M, Potter GG, McQuoid DR, Steffens DC, Beyer JL, Taylor WD (2017): Longitudinal Cognitive Outcomes of Clinical Phenotypes of Late-Life Depression. Am J Geriatr Psychiatry. 25:1123–1134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Castren E, Monteggia LM (2021): Brain-Derived Neurotrophic Factor Signaling in Depression and Antidepressant Action. Biol Psychiatry. 90:128–136. [DOI] [PubMed] [Google Scholar]
- 54.Collin SH, Milivojevic B, Doeller CF (2015): Memory hierarchies map onto the hippocampal long axis in humans. Nat Neurosci. 18:1562–1564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Videbech P, Ravnkilde B (2004): Hippocampal volume and depression: a meta-analysis of MRI studies. Am J Psychiatry. 161:1957–1966. [DOI] [PubMed] [Google Scholar]
- 56.Cole J, Toga AW, Hojatkashani C, Thompson P, Costafreda SG, Cleare AJ, et al. (2010): Subregional hippocampal deformations in major depressive disorder. J Affect Disord. 126:272–277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Mathias SR, Knowles EE, Kent JW Jr., McKay DR, Curran JE, de Almeida MA, et al. (2016): Recurrent major depression and right hippocampal volume: A bivariate linkage and association study. Hum Brain Mapp. 37:191–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Shi F, Liu B, Zhou Y, Yu C, Jiang T (2009): Hippocampal volume and asymmetry in mild cognitive impairment and Alzheimer's disease: Meta-analyses of MRI studies. Hippocampus. 19:1055–1064. [DOI] [PubMed] [Google Scholar]
- 59.Minkova L, Habich A, Peter J, Kaller CP, Eickhoff SB, Kloppel S (2017): Gray matter asymmetries in aging and neurodegeneration: A review and meta-analysis. Hum Brain Mapp. 38:5890–5904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Pusparani Y, Lin CY, Jan YK, Lin FY, Liau BY, Alex JSR, et al. (2025): Hippocampal volume asymmetry in Alzheimer disease: A systematic review and meta-analysis. Medicine (Baltimore). 104:e41662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Maass A, Berron D, Harrison TM, Adams JN, La Joie R, Baker S, et al. (2019): Alzheimer's pathology targets distinct memory networks in the ageing brain. Brain. 142:2492–2509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Steffens DC, Byrum CE, McQuoid DR, Greenberg DL, Payne ME, Blitchington TF, et al. (2000): Hippocampal volume in geriatric depression. Biol Psychiatry. 48:301–309. [DOI] [PubMed] [Google Scholar]
- 63.Andreescu C, Ajilore O, Aizenstein HJ, Albert K, Butters MA, Landman BA, et al. (2019): Disruption of Neural Homeostasis as a Model of Relapse and Recurrence in Late-Life Depression. Am J Geriatr Psychiatry. 27:1316–1330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Ballester PL, Suh JS, Nogovitsyn N, Hassel S, Strother SC, Arnott SR, et al. (2021): Accelerated brain aging in major depressive disorder and antidepressant treatment response: A CAN-BIND report. Neuroimage Clin. 32:102864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Loureiro JC, Stella F, Pais MV, Radanovic M, Canineu PR, Joaquim HPG, et al. (2020): Cognitive impairment in remitted late-life depression is not associated with Alzheimer's disease-related CSF biomarkers. J Affect Disord. 272:409–416. [DOI] [PubMed] [Google Scholar]
- 66.Mackin RS, Insel PS, Landau S, Bickford D, Morin R, Rhodes E, et al. (2021): Late-Life Depression Is Associated With Reduced Cortical Amyloid Burden: Findings From the Alzheimer's Disease Neuroimaging Initiative Depression Project. Biol Psychiatry. 89:757–765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67.Karim HT, Lee S, Gerlach A, Stinley M, Berta R, Mahbubani R, et al. (2024): Hippocampal subfield volume in older adults with and without mild cognitive impairment: Effects of worry and cognitive reappraisal. Neurobiol Aging. 141:55–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68.Monereo-Sanchez J, Jansen JFA, van Boxtel MPJ, Backes WH, Kohler S, Stehouwer CDA, et al. (2024): Association of hippocampal subfield volumes with prevalence, course and incidence of depressive symptoms: The Maastricht Study. Br J Psychiatry. 224:66–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69.Taylor WD, McQuoid DR, Payne ME, Zannas AS, MacFall JR, Steffens DC (2014): Hippocampus atrophy and the longitudinal course of late-life depression. Am J Geriatr Psychiatry. 22:1504–1512. [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.