Abstract
Background:
Neuropsychiatric symptoms (NPS) among cognitively normal older adults are increasingly recognized as risk factors for cognitive decline and impairment. However, the underlying mechanisms remain unclear.
Objective:
To examine whether biomarkers of Alzheimer’s disease (amyloid burden) and cerebrovascular disease (white matter hyperintensity (WMH) volume) modify the association between NPS and cognitive decline among cognitively unimpaired older adults.
Methods:
Analyses included 193 cognitively unimpaired participants (M age = 70 years) from the BIOCARD study, including 148 with PET amyloid and WMH biomarker data. NPS were measured with Neuropsychiatric Inventory and Geriatric Depression Scale scores. Linear mixed effects models were used to examine the association between baseline NPS and longitudinal cognitive trajectories (M follow-up = 3.05 years), using separate models for global, episodic memory, and executive function cognitive composite scores. In a subset of individuals with biomarker data, we evaluated whether WMH or cortical amyloid burden modified the relationship between NPS and cognitive change (as indicated by the NPS × biomarker × time interactions).
Results:
Higher baseline NPS were associated with lower executive function scores, but not a faster rate of decline in executive function. NPS symptoms were unrelated to the global or episodic memory composite scores, and there was little evidence of a relationship between NPS symptoms and cognitive change over time. The associations between NPS and cognitive decline did not differ by amyloid or WMH burden, and NPS were unrelated to amyloid and WMH burden.
Conclusion:
These results suggest that the effect of neuropsychiatric symptoms on executive dysfunction may occur through mechanisms outside of amyloid and cerebrovascular disease.
Keywords: Alzheimer’s disease, amyloid, depression, mild behavioral impairment, neuropsychiatric symptoms, white matter hyperintensities
INTRODUCTION
Neuropsychiatric symptoms (NPS) in late-life are increasingly recognized as risk factors for dementia and mild cognitive impairment (MCI), an early symptomatic precursor to Alzheimer’s disease (AD) [1]. NPS have been associated with progression to MCI/AD dementia [2–5] and more rapid cognitive decline in cognitively unimpaired older adults [6–8]. New NPS, even at low or subsyndromal severity, occurring in advance of or co-occurring with MCI, have been referred to as mild behavioral impairment [9]. Over 59% of individuals who are later diagnosed with MCI or dementia experience NPS prior to their diagnosis, with symptoms of depression and irritability being the most common NPS preceding the onset of MCI [10]. Although the relationship between NPS and cognitive decline has been demonstrated across multiple cohorts, the mechanisms underlying this association remain unclear.
Previous studies examining the association between NPS and amyloid in cognitively normal subjects have produced mixed results. Depressive symptoms have been investigated most extensively in this context, with some studies reporting an association between subsyndromal depressive symptoms and amyloid [11, 12], and others reporting no association [13–15] or mixed findings [16] in cognitively unimpaired older adults. Outside of depression, anxiety [11, 12] has also been associated with amyloid biomarkers, Studies that measured a wider variety of NPS using the Neuropsychiatric Inventory (NPI), however, have reported no cross-sectional relationship [13, 16], but some evidence of greater 1-year increase of NPS in individuals with higher biomarker levels [16].
Only a few prior studies have examined the interaction between NPS and AD biomarker measures on cognition in cognitively unimpaired older adults, and these studies have yielded mixed results. A cross-sectional study that investigated the interaction between subsyndromal depressive symptoms and AD biomarkers found that depressive symptoms were associated with worse episodic memory and global cognitive function (but not executive function, processing speed, or language) in individuals with evidence of both amyloid pathology and neurodegeneration [17]. To our knowledge, only one prior prospective longitudinal study has investigated this interaction in relation to longitudinal cognitive change in cognitively unimpaired older adults. Gatchel et al. reported that increasing depressive symptoms were associated with worsening global cognition over time in individuals with higher, but not lower cortical amyloid burden as measured by positron emission tomography (PET) [18]. However, a retrospective pathological study reported that, while symptoms of depression averaged over time were associated with more rapid global cognitive decline prior to death, there was no significant interaction with neuropathological markers of AD or cerebrovascular disease [8].
A possible explanation for the discrepancy across studies examining NPS and AD biomarkers may be, in part, attributable to the presence of mixed dementia pathologies. Etiologies outside of amyloid and tau, including vascular factors, have also been proposed as a possible contributor in the relationship between depressive symptoms and dementia [19]. For example, we recently found that subsyndromal depressive symptoms at middle age (mean age = 57 years) were associated with risk of progression from normal cognition to MCI in individuals with low levels of cerebrospinal fluid (CSF) amyloid and tau [20]. In exploratory analyses to examine the potential role of vascular risk factors as a non-AD mechanism underlying the association between depressive symptoms and progression to MCI/dementia, we observed that individuals with low CSF amyloid and tau who progressed to MCI/dementia had higher rates of vascular risk factors and white matter hyperintensity (WMH) volumes relative to those who remained normal. Taken together with observations that depressive symptoms are associated with increased risk of MCI symptom onset [21], our findings suggested that this relationship may be mediated through mechanisms independent of amyloid and tau. However, the analyses were primarily limited to baseline depressive symptoms and AD pathology measured at middle age, and the relationship between depressive symptoms and AD pathology on cognition among older adults was not examined.
The present study builds on existing literature in several ways. Prior studies on the interaction between NPS and neuropathology on longitudinal cognitive decline have focused on symptoms of depression and global cognitive measures. However, 1) NPS other than depression have been associated with risk of progression from normal cognition to MCI/AD [2, 3, 9, 22] and should therefore be examined in relation to biomarkers of AD and cerebrovascular disease; and 2) profiles of cognitive impairment due to vascular factors may differ from AD at early stages, with prior studies suggesting a stronger relationship with domains of executive function and processing speed [23, 24]. It therefore remains unknown whether NPS other than depression interact with neuropathology biomarkers in relation to cognitive decline, and whether these relationships differ for global versus domain-specific composite scores. To expand on our previous findings and explore these gaps in the literature, we examined 1) neuropsychiatric symptoms and depressive symptoms in relation to global and domain-specific cognitive changes; 2) if this relationship is modified by biomarkers of AD and small-vessel cerebrovascular disease; and 3) the cross-sectional relationship between biomarkers of AD, cerebrovascular disease, and neuropsychiatric symptoms.
METHODS
Study design
Data for these analyses were derived from the BIOCARD study. The BIOCARD study is an ongoing longitudinal study, initiated in 1995 at the National Institutes of Health (NIH) to identify predictors of progression from normal cognition to symptoms of AD. Recruitment procedures and baseline evaluations have previously been described in detail [25]. By design, 75% of participants had a first-degree relative with AD. At study baseline, individuals were excluded from participation if they were cognitively impaired or had significant medical problems such as severe cardiovascular disease (e.g., atrial fibrillation), chronic neurologic disorders (e.g., epilepsy, multiple sclerosis), or severe cerebrovascular disease based on magnetic resonance imaging (MRI) scan. After providing written informed consent, 349 individuals were enrolled in the study. The study was conducted at the NIH from 1995–2005. During this time, participants were administered a comprehensive clinical and neuropsychological examination annually. MRI scans, CSF samples, and blood specimens were obtained approximately every 2 years (Supplementary Figure 1). The study was stopped in 2005 for administrative reasons and was re-initiated at Johns Hopkins University (JHU) in 2009. Individuals who were recruited while the study was conducted at the NIH were re-enrolled and annual clinical and neuropsychiatric assessments, comprehensive neuropsychological assessments, and blood draws were resumed. In 2015, biennial collection of cerebrospinal fluid and MRI scans was re-established, and 11C-Pittsburgh Compound B (PiB) amyloid PET started.
Participant selection
The present report examined data from 193 cognitively unimpaired participants who had NPS and cognitive measures from 2015 (after PiB PET imaging was initiated) onwards (the “full” sample). Further analyses were completed in the “biomarker” sample, defined as the subset of individuals who also had an amyloid PET scan, as well as WMH data within 15 months of the amyloid PET scan (n = 148; mean gap between PET and MRI scan = 2.5 (SD = 29.2) days). For these analyses, ‘baseline’ was defined as time of first PiB PET scan (for the biomarker sample), or first visit from 2015 with both NPS and cognitive data available (for those without biomarkers). All data were collected between January 2015 and May 2020.
Neuropsychiatric symptom assessments
Since the study has been at JHU, neuropsychiatric symptoms have been measured with the 15-item Geriatric Depression Scale (GDS) and the 12-item NPI. The GDS is a self-report screening tool designed to assess depressive symptoms in older adults. A higher score indicates higher severity of depressive symptoms, and scores under 5 are considered to be normal [26]. GDS scores were treated as both a continuous measure and, given the low mean symptom severity, alternatively dichotomized at 0 versus ≥ 1.
The NPI is a structured informant-based interview that assesses the following 12 neuropsychiatric symptom domains: delusions, hallucinations, agitation or aggression, depression or dysphoria, anxiety, elation or euphoria, apathy or indifference, disinhibition, irritability or lability, appetite and eating, aberrant motor behavior, and nighttime behaviors [27]. All NPI were completed by a study partner who knows the individual well (i.e., usually a spouse, child, sibling, or good friend). One point was given for the presence of symptoms in each domain during the past month. NPI scores were treated as both a continuous measure and dichotomized at 0 versus ≥ 1 for similar reasons as described above for the GDS.
Clinical and cognitive assessments
A consensus diagnosis for each study visit was established by the staff of the BIOCARD Clinical Core at Johns Hopkins. The diagnostic criteria followed the recommendations incorporated in the National Institute on Aging and the Alzheimer’s Association working group reports for the diagnosis of MCI [28] and dementia due to AD [29]. Consensus diagnosis procedures have previously been described in detail elsewhere [25]. Briefly, a syndromic diagnosis (i.e., cognitively normal, MCI, impaired not MCI, or dementia) is established using 1) clinical data pertaining to the subject’s medical, neurological, and psychiatric status; 2) reports of changes in cognition by the subject and collateral sources (based on the Clinical Dementia Rating Scale (CDR) [30]); and 3) performance on neuropsychological tests longitudinally and in comparison to published norms. Individuals with a diagnosis of “impaired not MCI” at baseline were also included in the “normal” group as they did not meet criteria for MCI.
A comprehensive neuropsychological battery was completed annually (see [25] for the complete battery). For the present study, we considered three cognitive composite scores. A global cognitive composite score was derived from four tests previously determined to provide the best combination of cognitive predictors of progression from normal cognition to MCI in the BIOCARD cohort [25]: Paired Associates Immediate Recall (Wechsler Memory Scale-Revised; WMS-R), Logical Memory Delayed Recall (Story A; WMS-R), Boston Naming, and Digit-Symbol Substitution (Wechsler Adult Intelligence Scale-Revised; WAIS-R). To calculate the composite score, the four measures were converted to z-scores (based on means and SDs from all BIOCARD participants’ first visit at JHU), and then averaged, with the requirement that at least 2 of 4 scores were present at a given time point.
Domain-specific cognitive composite scores were also calculated for verbal episodic memory (including Paired Associates immediate recall, Logical Memory delayed recall, and California Verbal Learning Test total recall over trials 1–5) and executive function (including Digit-Symbol Substitution, Digit Span backwards (WMS-R), and Trail Making Test part B), using factor analytic techniques described previously [31]. For each cognitive domain, standardized task scores were weighted by their factor loadings, then averaged to generate a composite score. Cognitive trajectories over time were calculated from ‘baseline’ (as defined above) and included all available follow-up.
White matter hyperintensity volumes
WMH volumes were derived from axial fluid-attenuated inversion recovery (FLAIR) images obtained on a 3T MR system (Philips Healthcare, Best, The Netherlands) as part of a multi-modal MRI protocol. A previously described, an automated method [32, 33] was used to quantify global WMH volumes. Because the distribution of baseline WMH volumes was skewed, all analyses were run using a continuous log(x + 1)-transformed WMH variable and a dichotomized WMH variable (1 = highest quartile considering WMH data across all time points, 0 = lower three quartiles).
Cortical PiB DVR
Dynamic 11C-PiB PET scans were obtained on 3D mode from a GE Advance scanner, started immediately after intravenous bolus injection of 11C-PiB. Distribution volume ratio (DVR) images were computed using a simplified reference tissue model [34] with cerebellar gray matter as the reference region. Mean cortical amyloid-β (Aβ) burden was calculated as the average of the DVR values in anterior, middle, and posterior cingulate regions, superior, middle and inferior frontal and orbitofrontal, superior parietal, supramarginal and angular gyrus regions, superior, middle and inferior temporal, pre-cuneus, superior, middle and inferior occipital, excluding the sensorimotor strip. All analyses were run using PiB cortical DVR (cDVR) both as a continuous variable and a dichotomized variable based on PiB positivity (defined as cDVR > 1.061 based on 2-class Gaussian mixture modeling [35]).
Statistical analysis
Group differences in baseline characteristics of participants were assessed with Wilcoxon rank-sum test for continuous variables or chi-square tests for dichotomous variables.
The first set of models were run in the full sample, and used linear mixed effects models to examine the association between baseline NPS and longitudinal cognitive trajectories. Separate models were run for the global, episodic memory, and executive function cognitive composite scores. Covariates in the models included the baseline NPS score, and terms for baseline age, sex, years of education, and the interaction of each predictor with time. Running this analysis with the full sample allowed us to examine the main effect of NPS on cognitive trajectories with a larger sample size to increase statistical power to detect such associations.
In the subset of individuals with FLAIR and PiB PET scans (i.e., the biomarker sample), we re-ran the above models, this time including additional terms for baseline WMH volume or baseline amyloid PET, and their interaction with time, as well as the three-way NPS × biomarker × time interactions (i.e., NPS × WMH volume × time or NPS × amyloid PET × time) in order to evaluate whether baseline WMH or amyloid PET modified the relationship between NPS and subsequent cognitive change. If the three-way interaction was not significant, we ran reduced models that excluded the 3-way interaction term as well as the NPS × biomarker lower order interaction term (i.e., NPS × WMH or NPS × amyloid PET).
Lastly, linear and logistic regression models were used to assess the cross-sectional association between NPS (outcome) and biomarker levels, covarying age, sex and education. Given the limited existing literature on the interaction between NPS and AD biomarkers on cognitive change, we did not correct for multiple comparisons to allow for hypothesis generation in this exploratory study [36].
For all analyses, the significance level was set at p < 0.05. All analyses were run in R (version 4.0.3).
RESULTS
Out of the 349 individuals that were originally enrolled, 148 met inclusion criteria for the biomarker sample. Numbers of subjects excluded and reasons for exclusion included: 28 due to withdrawal from study/not re-enrolled, 153 had no PiB PET scan, 4 did not have WMH imaging within 15 months of their PiB PET scan, 16 did not have normal cognition at baseline. 45 individuals who had visits with both NPS and cognitive measures between 2015 and 2019 and normal cognition at baseline, but no PiB PET and WMH imaging, were added to the biomarker sample to form the full sample. Out of the full sample (N = 193), 25 subjects had one gap of > 18 months between consecutive cognitive assessment visits. 1 subject had two gaps of > 18 months between consecutive cognitive visits. There were no missing data for model covariates including age, sex, and education.
Baseline characteristics for participants are shown in Table 1, for the full sample and for the subset with PiB and WMH imaging (biomarker sample). Participants in the full sample had an average of 3.82 visits (SD = 1.3, range 1–6) over 3.05 years of follow-up (SD = 1.2, range 0–5.04). Baseline MRI scans were completed on average 2.5 (SD = 29.2) days after PiB PET scan was completed. For participants in the biomarker sample, baseline cognitive assessment and NPS assessment were both completed on the same day as PiB PET scan.
Table 1.
Descriptive statistics for participants
| Full sample | Biomarker sample | |
|---|---|---|
| N | 193 | 148 |
| Age, Mean (SD) | 70.03 (8.59) | 69.22 (8.43) |
| Female sex, N (%) | 122 (63.21%) | 92 (62.16%) |
| White race, N (%) | 190 (98.45%) | 145 (97.97%) |
| Years of education, Mean (SD) | 17.32 (2.28) | 17.39 (2.23) |
| Global cognitive composite score, Mean (SD) | 0.46 (0.54) | 0.51 (0.52) |
| Episodic memory composite, Mean (SD) | 0.48 (0.46) | 0.53 (0.46) |
| Executive function composite, Mean (SD) | 0.14 (0.41) | 0.22 (0.42) |
| Baseline GDS total score, Mean (SD) | 0.99 (1.59) | 0.86 (1.52) |
| GDS ≥ 1, N (%) | 90 (46.63%) | 61 (41.22%) |
| GDS ≥ 2, N (%) | 42 (21.76%) | 28 (18.92%) |
| GDS ≥ 5, N (%) | 9 (4.66%) | 5 (3.38%) |
| Baseline NPI total score, Mean (SD) | 0.29 (0.72) | 0.29 (0.73) |
| Baseline NPI 0 versus 1+, N (%) | 35 (18.13%) | 27 (18.24%) |
| Number of visits over time, Mean (SD) | 3.82 (1.26) | 4.01 (1.20) |
| Years between baseline and last visit, Mean (SD) | 3.05 (1.24) | 3.17 (1.20) |
| PiB cortical DVR, Mean (SD) | – | 1.08 (0.15) |
| PiB Positive, N (%) | – | 44 (30.56%) |
| WMH volume cm3, Mean (SD) | – | 4399.83 (7099.06) |
| WMH volume in upper quartile, N (%) | – | 27 (18.24%) |
Biomarker sample, Individuals with WMH scan and 11C-PiB PET scan; N, number; SD, standard deviation; GDS, Geriatric Depression Scale-15; NPI, Neuropsychiatric Inventory; PiB, 11C-Pittsburgh Compound B; DVR, Distribution volume ratio; WMH, white matter hyperintensities.
Average severity of neuropsychiatric symptoms was low. The mean GDS score was 0.99 (SD = 1.59), well below the generally used cut-off for depression of ≥ 5 [26]. The mean NPI total score was 0.29 (SD = 0.72) and 35 (18.13%) individuals had NPI total score ≥ 1. The most frequently reported NPI symptoms were irritability or lability (n = 16, 8.29%) and depression (n = 11, 5.70%) (Table 2). In individuals who also had WMH and cortical PiB measures, the patterns were similar.
Table 2.
Most commonly reported NPI symptoms in individuals with NPI total scores ≥ 1
| Full sample N = 193 |
Biomarker sample N = 148 |
|
|---|---|---|
| NPI variable | N (%) | N (%) |
| ≥ 1 NPI total score | 35 (18.13%) | 27 (18.24%) |
| Delusions | 0 (0%) | 0 (0%) |
| Hallucinations | 0 (0%) | 0 (0%) |
| Agitation or Aggression | 8 (4.15%) | 6 (4.05%) |
| Depression or Dysphoria | 11 (5.70%) | 8 (5.41%) |
| Anxiety | 8 (4.15%) | 6 (4.05%) |
| Elation or Euphoria | 0 (0%) | 0 (0%) |
| Apathy or Indifference | 4 (2.07%) | 3 (2.03%) |
| Disinhibition | 2 (1.04%) | 2 (1.35%) |
| Irritability or Lability | 16 (8.29%) | 13 (8.78%) |
| Motor disturbance | 0 (0%) | 0 (0%) |
| Nighttime behaviors | 5 (2.59%) | 4 (2.7%) |
| Appetite and Eating | 2 (1.04%) | 1 (0.68%) |
Biomarker sample, Individuals with WMH scan and 11C-PiB PET scan; N, number.
Results of the linear mixed effects models examining the relationship between NPS and cognition in the “full” sample are shown in Table 3. For most models, there were significant main effects of age (indicating lower levels of cognitive performance with older baseline age) and sex (indicating higher levels of cognitive performance among females), as well as significant age × time interactions (indicating greater rates of cognitive decline with older baseline ages). Higher baseline GDS (dichotomous) and NPI (continuous and dichotomous) symptoms were associated with lower executive function scores (all p < 0.04), but were unrelated to global cognition or episodic memory (all p > 0.20). With one exception (dichotomous GDS scores), neuropsychiatric symptoms were not associated with short-term cognitive trajectories in the global or domain-specific composite scores (all p > 0.08). GDS scores ≥ 1 were weakly associated with less decline in executive function composite scores relative to GDS scores = 0 (p = 0.041).
Table 3.
Longitudinal mixed effects model results for baseline neuropsychiatric symptoms (NPS) in relation to the global cognitive composite score, episodic memory, and executive function (full sample; N = 193)
| Global cognition |
Episodic memory |
Executive function |
||||
|---|---|---|---|---|---|---|
| Estimate (SE) | p | Estimate (SE) | p | Estimate (SE) | p | |
| GDS total | ||||||
| Time | −0.012 (0.024) | 0.615 | 0.028 (0.025) | 0.269 | −0.093 (0.012) | 0.001 |
| GDS total | 0.052 (0.040) | 0.198 | 0.05 (0.042) | 0.232 | −0.042 (0.041) | 0.31 |
| GDS total × time | −0.007 (0.008) | 0.404 | −0.005 (0.008) | 0.511 | 0.006 (0.007) | 0.413 |
| GDS ≥ 1 | ||||||
| Time | −0.018 (0.027) | 0.508 | 0.013 (0.028) | 0.639 | −0.114 (0.023) | <0.001 |
| GDS ≥ 1 | 0.020 (0.129) | 0.875 | 0.040 (0.134) | 0.766 | −0.267 (0.131) | 0.043 |
| GDS ≥ 1 × time | −0.004 (0.027) | 0.888 | 0.014 (0.028) | 0.609 | 0.048 (0.023) | 0.041 |
| NPI total | ||||||
| Time | −0.018 (0.023) | 0.421 | 0.028 (0.024) | 0.247 | −0.087 (0.020) | <0.001 |
| NPI total | −0.003 (0.089) | 0.969 | 0.046 (0.093) | 0.621 | −0.267 (0.090) | 0.003 |
| NPI total × time | −0.005 (0.020) | 0.804 | −0.022 (0.021) | 0.307 | 0.002 (0.018) | 0.921 |
| NPI ≥ 1 | ||||||
| Time | −0.016 (0.023) | 0.486 | 0.032 (0.024) | 0.179 | −0.087 (0.020) | <0.001 |
| NPI ≥ 1 | −0.039 (0.166) | 0.817 | 0.043 (0.173) | 0.806 | −0.457 (0.168) | 0.007 |
| NPI ≥ 1 × time | −0.024 (0.035) | 0.497 | −0.066 (0.037) | 0.078 | 0.007 (0.031) | 0.835 |
All models adjusted for baseline age, gender, education, and their interactions with time.
The pattern of results was similar when GDS was dichotomized as ≤ 1 versus ≥ 2 (Supplementary Table 1), except that the dichotomous GDS × time interaction for executive function was no longer significant. When the same model was run with the biomarker sample only (n = 148), higher baseline NPI (continuous) remained associated with lower executive function scores (p = 0.045), while dichotomous NPI trended towards lower executive function scores (Supplementary Table 2).
Results of the models examining the relationship between NPS symptoms, biomarkers (dichotomous WMH volume or PiB positivity) and cognition are shown in Tables 4 and 5, respectively. Sensitivity analysis using GDS ≤ 1 versus ≥ 2 are shown in Supplementary Tables 3 and 4. There was no effect on 3-way NPS × biomarker × time interactions, and no NPS × biomarker interactions (all p > 0. 05; data not shown), indicating that the relationship between NPS and cognition was not modified by biomarker levels. In the reduced models, higher NPI (continuous) symptoms were associated with lower executive function scores after accounting for WMH (p = 0.049) or PiB positivity (p = 0.038). WMH volume was associated with lower levels of global cognition and episodic memory, but not executive function, while PiB positivity was not associated with any cognitive measures. WMH was not associated with change in cognition over time. There was a trend towards an association between PiB positivity and greater rate of decline in executive function composite score over time. However, the association only reached significance for GDS ≥ 1. The patterns of results were similar in the models using continuous biomarker levels (i.e., log-transformed WMH volume and PiB cDVR; see Supplementary Tables 5 and 6).
Table 4.
Longitudinal mixed effects model results for baseline neuropsychiatric symptoms (NPS), and WMH volumes, in relation to the global cognitive composite score, episodic memory and executive function (biomarker sample; N = 148)
| WMH volume - Dichotomous | ||||||
|---|---|---|---|---|---|---|
|
| ||||||
| Global cognition |
Episodic memory |
Executive function |
||||
| Estimate (SE) | p | Estimate (SE) | p | Estimate (SE) | p | |
| GDS total | ||||||
| Time | −0.023 (0.027) | 0.405 | 0.042 (0.031) | 0.185 | −0.099 (0.025) | <0.001 |
| GDS total | 0.068 (0.046) | 0.145 | 0.040 (0.047) | 0.393 | 0.026 (0.049) | 0.602 |
| WMH volume | −0.465 (0.195) | 0.019 | −0.581 (0.197) | 0.004 | −0.135 (0.206) | 0.515 |
| GDS Total × time | −0.005 (0.01) | 0.621 | −0.007 (0.011) | 0.524 | −0.002 (0.009) | 0.829 |
| WMH volume × time | −0.067 (0.043) | 0.126 | −0.076 (0.050) | 0.134 | −0.012 (0.040) | 0.756 |
| GDS ≥ 1 | ||||||
| Time | −0.040 (0.03) | 0.182 | 0.022 (0.034) | 0.522 | −0.117 (0.027) | <0.001 |
| GDS ≥ 1 | 0.022 (0.144) | 0.876 | 0.034 (0.144) | 0.812 | −0.132 (0.151) | 0.384 |
| WMH volume | −0.476 (0.197) | 0.017 | −0.587 (0.197) | 0.003 | −0.139 (0.206) | 0.500 |
| GDS ≥ 1 × time | 0.024 (0.030) | 0.429 | 0.024 (0.035) | 0.498 | 0.031 (0.028) | 0.274 |
| WMH volume × time | −0.067 (0.043) | 0.123 | −0.076 (0.050) | 0.131 | −0.013 (0.040) | 0.749 |
| NPI total | ||||||
| Time | −0.029 (0.026) | 0.274 | 0.040 (0.030) | 0.183 | −0.100 (0.024) | <0.001 |
| NPI total | −0.030 (0.098) | 0.759 | 0.030 (0.098) | 0.761 | −0.200 (0.101) | 0.049 |
| WMH volume | −0.474 (0.197) | 0.017 | −0.591 (0.197) | 0.003 | −0.117 (0.204) | 0.568 |
| NPI total × time | 0.004 (0.024) | 0.862 | −0.022 (0.028) | 0.438 | −0.005 (0.023) | 0.837 |
| WMH volume × time | −0.067 (0.043) | 0.125 | −0.072 (0.050) | 0.156 | −0.011 (0.040) | 0.777 |
| NPI ≥ 1 | ||||||
| Time | −0.026 (0.026) | 0.319 | 0.045 (0.030) | 0.135 | −0.101 (0.020) | <0.001 |
| NPI ≥ 1 | 0.011 (0.185) | 0.952 | 0.120 (0.185) | 0.519 | −0.314 (0.190) | 0.106 |
| WMH volume | −0.479 (0.199) | 0.017 | −0.606 (0.199) | 0.003 | −0.088 (0.207) | 0.670 |
| NPI ≥ 1 × time | −0.010 (0.042) | 0.806 | −0.071 (0.048) | 0.141 | −0.004 (0.039) | 0.927 |
| WMH volume × time | −0.064 (0.044) | 0.147 | −0.061 (0.051) | 0.233 | −0.011 (0.041) | 0.782 |
All models adjusted for baseline age, gender, education, and their interactions with time. All two- and three-way interactions of NPS × WMH dichotomous, and NPS × WMH dichotomous × time were not significant (all p > 0. 05) and therefore excluded from the final models. WMH positive, whole brain WMH volume in highest quartile.
Table 5.
Longitudinal mixed effects model results for baseline neuropsychiatric symptoms (NPS), and PiB positivity, in relation to the global cognitive composite score, episodic memory and executive function (biomarker sample; N = 148)
| PiB positive | ||||||
|---|---|---|---|---|---|---|
|
| ||||||
| Global cognition |
Episodic memory |
Executive function |
||||
| Estimate (SE) | p | Estimate (SE) | p | Estimate (SE) | p | |
| GDS total | ||||||
| Time | −0.022 (0.028) | 0.430 | 0.034 (0.031) | 0.279 | −0.087 (0.025) | 0.001 |
| GDS total | 0.072 (0.047) | 0.127 | 0.045 (0.048) | 0.349 | 0.032 (0.049) | 0.516 |
| Amyloid | 0.019 (0.157) | 0.905 | −0.098 (0.161) | 0.546 | 0.089 (0.164) | 0.587 |
| GDS total × time | −0.004 (0.010) | 0.692 | −0.005 (0.011) | 0.638 | −0.002 (0.009) | 0.806 |
| Amyloid × time | −0.041 (0.034) | 0.235 | −0.025 (0.038) | 0.513 | −0.062 (0.031) | 0.052 |
| GDS ≥ 1 | ||||||
| Time | −0.04 (0.030) | 0.194 | 0.012 (0.034) | 0.734 | −0.103 (0.028) | <0.001 |
| GDS ≥ 1 | 0.053 (0.148) | 0.722 | 0.051 (0.151) | 0.735 | −0.091 (0.153) | 0.556 |
| Amyloid | 0.03 (0.158) | 0.852 | −0.092 (0.161) | 0.569 | 0.101 (0.164) | 0.539 |
| GDS ≥ 1 × time | 0.027 (0.031) | 0.382 | 0.032 (0.034) | 0.351 | 0.026 (0.028) | 0.358 |
| Amyloid × time | −0.042 (0.034) | 0.223 | −0.026 (0.038) | 0.490 | −0.062 (0.031) | 0.048 |
| NPI total | ||||||
| Time | −0.026 (0.027) | 0.336 | 0.036 (0.030) | 0.228 | −0.089 (0.024) | <0.001 |
| NPI total | −0.053 (0.099) | 0.593 | 0.015 (0.101) | 0.885 | −0.212 (0.101) | 0.038 |
| Amyloid | 0.04 (0.158) | 0.802 | −0.091 (0.162) | 0.573 | 0.122 (0.162) | 0.451 |
| NPI total × time | 0.000 (0.024) | 0.992 | −0.029 (0.027) | 0.283 | 0.001 (0.022) | 0.963 |
| Amyloid × time | −0.041 (0.034) | 0.232 | −0.021 (0.038) | 0.581 | −0.062 (0.031) | 0.050 |
| NPI ≥ 1 | ||||||
| Time | −0.022 (0.027) | 0.411 | 0.042 (0.030) | 0.154 | −0.090 (0.024) | <0.001 |
| NPI ≥ 1 | −0.071 (0.816) | 0.705 | 0.040 (0.190) | 0.832 | −0.341 (0.191) | 0.076 |
| Amyloid | 0.038 (0.159) | 0.811 | −0.093 (0.162) | 0.565 | 0.121 (0.163) | 0.459 |
| NPI ≥ 1 × time | −0.025 (0.041) | 0.546 | −0.091 (0.046) | 0.049 | 0.006 (0.038) | 0.868 |
| Amyloid × time | −0.038 (0.034) | 0.265 | −0.016 (0.038) | 0.670 | −0.063 (0.032) | 0.050 |
All models adjusted for baseline age, gender, education, and their interactions with time. All two- and three-way interactions of NPS × PiB positivity, and NPS × PiB positivity × time were not significant (all p > 0.05) and therefore excluded from the final models. PiB positive, mean PiB cortical DVR > 1.061.
Cross-sectionally, baseline NPS were not associated with WMH volumes or amyloid PET (all p > 0.45; see Supplementary Table 7 for the results of the continuous and dichotomous models).
DISCUSSION
In this analysis of cognitively unimpaired older adults with low levels of neuropsychiatric symptoms, we observed a significant association between higher neuropsychiatric symptoms and lower level of executive functioning. The relationship between NPI-total and executive function was independent of baseline amyloid burden or WMH volume. In contrast, baseline NPS were not associated with levels of global cognition or episodic memory performance, or with cognitive trajectories in any domain. NPS were also not associated with PET amyloid burden or small vessel cerebrovascular disease, as measured by WMH volume. Of note, the association between depressive symptoms and executive function was not significant when biomarkers were included in the model.
Consistent with a prior study on the time course of NPS prior to onset of MCI [10], we observed that the most common NPS were irritability or lability and depression or dysthymia. The association between neuropsychiatric symptoms and executive function is consistent with prior observations that executive dysfunction is common in depression, particularly in older adults [37]. While prior studies on the relationship between subsyndromal neuropsychiatric symptoms have reported a significant association with risk of progression to MCI/AD [2–5, 21], studies examining the relationship between subsyndromal neuropsychiatric symptoms and domain-specific cognitive decline among cognitively normal individuals are limited. To our knowledge, only one prior study has reported findings on baseline neuropsychiatric symptoms and cognitive decline across different cognitive domains in individuals who were cognitively normal at baseline, though the authors did not consider biomarker levels in their analyses. In this study of cognitively normal adults followed over 5.73 years with a mean NPI score of 0.26, higher NPI-total at baseline was associated with a more rapid rate of decline in recall and verbal fluency (but not in global cognition as measured by the Mini-Mental Status Examination or executive function as measured by the Symbol Digit Modalities Test), though they noted that the effect size of their findings were small [6]. When individual NPI symptoms were examined, NPI-depression was not associated with decline in any domain, and NPI-anxiety was associated with decline in divided attention only. Possible reasons for conflicting results between Burhanullah et al. and our study include our smaller sample size, our shorter length of follow-up of 3.05 years, and differences in cognitive assessments used. While, similar to Burhanullah et al., our analyses did not find evidence of an association between NPI and rate of global cognitive decline, GDS scores ≥ 1 were associated with slower executive function decline. This pattern, unexpected in view of evidence that depressive symptoms are associated with progression to cognitive impairment [21, 38, 39], may reflect a slower rate of decline in the context of lower baseline executive functioning in individuals with higher GDS scores. It should also be noted that this finding was not significant when GDS was considered as a continuous variable, or in sensitivity analysis using GDS ≤ 1 versus ≥ 2, suggesting the association between GDS and longitudinal change in executive function scores may be weak.
Studies examining the interaction between NPS and AD pathology on cognitive decline in individuals without cognitive impairment are limited, and have primarily focused on depressive symptoms. Similar to Wilson et al. (2014) [8], a retrospective neuropathological study of 1764 individuals without cognitive impairment at baseline followed for an average of 7.8 years, we did not find an interaction between NPS and AD- or vascular-related pathology on cognitive decline. However, other studies have reported contrasting findings. The most comparable study to our analyses is a longitudinal study of individuals from the Harvard Aging Brain Study (HABS), followed for an average of 4.4 years. This study found that increasing depressive symptoms (measured by change in GDS scores over time) were associated with worsening cognition over time (as measured by a multi-domain cognitive composite score) in individuals with higher, but not lower cortical amyloid burden (PiB DVR) [18]. It should be noted that, while both the HABS and BIOCARD global composite scores included a large contribution from memory tests, BIOCARD included a language test (Boston Naming) while HABS did not. These differences in analytic strategies, use of longitudinal as opposed to baseline GDS, and the structure of the global cognitive scores may, in part, account for differences in findings.
A similar interaction between NPS and AD biomarkers with respect to cognitive performance was reported by a cross-sectional study. Specifically, depressive symptoms were associated with worse episodic memory and global cognitive function (but not executive function) among individuals with evidence of both amyloid pathology and neurodegeneration, but not in those with only amyloid positivity or those without biomarker evidence AD pathology [17]. Although the cross-sectional nature of this study did not allow for investigation of cognitive change, their findings provide support that AD pathology, at later stages of preclinical AD [40], may have an interactive effect with depressive symptoms on cognitive performance. In contrast, our analyses, which did not consider neurodegeneration, did not find a significant interaction between amyloid (or WMH) and NPS on cognition.
Of note, we observed an association between NPS and executive function. This observation was consistent across different measures of NPS when biomarker status was not considered, but was only significant for the continuous NPI measure when biomarkers were added to the model, which may suggest that the association is weak. Similar to other studies discussed, the cognitive battery used in Javaherian et al. differed from ours, which may contribute to differences in findings. For example, their executive function battery additionally included Block Design from the Wechsler Intelligence Scale and Trail Making Test part A, which may additionally test components of visuospatial ability and visual attention.
Neuropsychiatric symptoms, in particular depression, have also been studied extensively in the context of cerebrovascular disease. Both depression and subsyndromal depressive symptoms have been associated with an elevated risk of vascular dementia [41, 42]. However, studies on the relationship between NPS and WMH in cognitively normal individuals have been limited. A longitudinal study using data from the Alzheimer’s Disease Neuroimaging Initiative reported that baseline WMH burden was associated with a greater increase in NPI-total scores over time only in MCI and AD, but not in cognitively unimpaired individuals [43]. Studies of individuals with cognitive impairment that have examined individual NPI domains and region-specific WMH have observed associations between specific neuropsychiatric symptoms and frontal WMH [44–46], though findings from analyses using total WMH have been mixed [46–48]. Future studies need to examine potential interactions between NPI and regional WMH and AD biomarker levels in relation to cognitive decline.
Our primary finding that the relationship between NPS and cognitive change over time is not modified by amyloid burden are consistent with our previous observations that subsyndromal depressive symptoms in middle age are associated with progression to MCI in individuals lacking CSF biomarker evidence of AD pathology, suggesting that the added risk of cognitive decline due to depression may be mediated through non-AD mechanisms [20]. Although we previously found in exploratory analyses that individuals with low AD pathology who progressed to MCI were more likely to have vascular risk factors and higher WMH volume at middle-age, our current findings did not provide evidence that neuropsychiatric symptoms are associated with short-term cognitive change, or that cerebrovascular disease burden increases the effect of neuropsychiatric symptoms on cognition. The differences in findings may reflect 1) the low number of participants included in our previous exploratory analyses (n = 32) and 2) that depressive symptoms at middle-age are associated with onset of MCI (over an average follow up of 12.7 years in our previous analyses), but not shorter-term cognitive change in late-life. Cognitive changes are subtle in preclinical stages of AD; for example, a longitudinal study of 1120 cognitively unimpaired individuals (mean age = 73 years) with evidence of amyloid pathology found that after 6 years of follow-up, participants began to experience cognitive impairment only approaching levels of early MCI [49]. Thus, the follow up period of 3 years may have been too short to detect a significant effect of NPS on cognitive decline in our cognitively unimpaired cohort, of which only 31% were amyloid-positive. Alternatively, it is also possible that the relationship between neuropsychiatric symptoms and cognitive change may follow a pattern similar to that observed between vascular risk factors and cognition, where vascular risk factors in middle-life, but not in late-life, are associated with increased risk of dementia [50, 51]. Future studies examining other potential mechanisms for the relationship between late-life neuropsychiatric symptoms and cognition outside of amyloid and vascular factors, such as tau, inflammation, neutrophin deficiency, impairments in neurotransmitter systems [19], and TDP-43 proteinopathy [52] are needed.
The study has several strengths, including a relatively large number of well-characterized, cognitively normal individuals with longitudinal follow up. However, our findings should be interpreted within the context of their limitations. Participants were well-educated and primarily Caucasian. They had relatively low prevalence and severity of neuropsychiatric symptoms, and the longitudinal follow-up included here was relatively short in duration (mean = 3.82 visits over a mean of 3.05 years). The low prevalence of neuropsychiatric symptomatology in the cohort may limit our power to detect smaller effect sizes and our findings may not be generalizable to individuals with more significant NPS. Furthermore, due to the low number of symptoms, we did not assess subscores of the GDS or NPI, which limits our understanding of possible differential effects of specific NPS on cognitive outcome. Additionally, the NPI was designed for use in individuals with dementia, who tend to have higher levels of neuropsychiatric symptoms than cognitively normal older adults, thus and may therefore be biased towards null results when used in cognitively normal individuals. Lastly, our analyses were not corrected for multiple comparisons. In the context of the relatively mixed literature, future studies with larger sample sizes and greater variability in NPI are needed to further examine these questions.
In summary, low-severity neuropsychiatric symptoms in older adults were associated with worse executive function cross sectionally, but not NPS were unrelated to short-term cognitive decline. This pattern of results was not modified by amyloid or WMH burden. These results suggest that the effect of neuropsychiatric symptoms on executive dysfunction may occur through mechanisms outside of amyloid and cerebrovascular disease.
Supplementary Material
ACKNOWLEDGMENTS
This work was supported by the National Institutes of Health [grant numbers U19-AG033655, P30 AG066507].
Footnotes
Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/21-5267r2).
SUPPLEMENTARY MATERIAL
The supplementary material is available in the electronic version of this article: https://dx.doi.org/10.3233/JAD-215267.
REFERENCES
- [1].Gauthier S, Reisberg B, Zaudig M, Petersen RC, Ritchie K, Broich K, Belleville S, Brodaty H, Bennett D, Chertkow H, Cummings JL, de Leon M, Feldman H, Ganguli M, Hampel H, Scheltens P, Tierney MC, Whitehouse P, Winblad B (2006) Mild cognitive impairment. Lancet 367, 1262–1270. [DOI] [PubMed] [Google Scholar]
- [2].Banks SJ, Raman R, He F, Salmon DP, Ferris S, Aisen P, Cummings J (2014) The Alzheimer’s disease cooperative study prevention instrument project: Longitudinal outcome of behavioral measures as predictors of cognitive decline. Dement Geriatr Cogn Dis Extra 4, 509–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Geda YE, Roberts RO, Mielke MM, Knopman DS, Christianson TJH, Pankratz VS, Boeve BF, Sochor O, Tangalos EG, Petersen RC, Rocca WA (2014) Baseline neuropsychiatric symptoms and the risk of incident mild cognitive impairment: A population-based study. AJP 171, 572–581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Leoutsakos JS, Forrester SN, Corcoran CD, Norton MC, Rabins PV, Steinberg MI, Tschanz JT, Lyketsos CG (2015) Latent classes of course in Alzheimer’s disease and predictors: The Cache County Dementia Progression Study. Int J Geriatr Psychiatry 30, 824–832. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Masters MC, Morris JC, Roe CM (2015) “Noncognitive” symptoms of early Alzheimer disease: A longitudinal analysis. Neurology 84, 617–622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Burhanullah MH, Tschanz JT, Peters ME, Leoutsakos JM, Matyi J, Lyketsos CG, Nowrangi MA, Rosenberg PB (2020) Neuropsychiatric symptoms as risk factors for cognitive decline in clinically normal older adults: The Cache County Study. Am J Geriatr Psychiatry 28, 64–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Köhler S, Allardyce J, Verhey FRJ, McKeith IG, Matthews F, Brayne C, Savva GM, Medical Research Council Cognitive Function and Ageing Study (2013) Cognitive decline and dementia risk in older adults with psychotic symptoms: A prospective cohort study. Am J Geriatr Psychiatry 21, 119–128. [DOI] [PubMed] [Google Scholar]
- [8].Wilson RS, Capuano AW, Boyle PA, Hoganson GM, Hizel LP, Shah RC, Nag S, Schneider JA, Arnold SE, Bennett DA (2014) Clinical-pathologic study of depressive symptoms and cognitive decline in old age. Neurology 83, 702–709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Ismail Z, Smith EE, Geda Y, Sultzer D, Brodaty H, Smith G, Agüera-Ortiz L, Sweet R, Miller D, Lyketsos CG, ISTAART Neuropsychiatric Symptoms Professional Interest Area (2016) Neuropsychiatric symptoms as early manifestations of emergent dementia: Provisional diagnostic criteria for mild behavioral impairment. Alzheimers Dement 12, 195–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Wise EA, Rosenberg PB, Lyketsos CG, Leoutsakos J-M (2019) Time course of neuropsychiatric symptoms and cognitive diagnosis in National Alzheimer’s Coordinating Centers volunteers. Alzheimers Dement (Amst) 11, 333–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Donovan NJ, Locascio JJ, Marshall GA, Gatchel J, Hanseeuw BJ, Rentz DM, Johnson KA, Sperling RA, Harvard Aging Brain Study (2018) Longitudinal association of amyloid beta and anxious-depressive symptoms in cognitively normal older adults. Am J Psychiatry 175, 530–537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Krell-Roesch J, Lowe VJ, Neureiter J, Pink A, Roberts RO, Mielke MM, Vemuri P, Stokin GB, Christianson TJ, Jack CR, Knopman DS, Boeve BF, Kremers WK, Petersen RC, Geda YE (2018) Depressive and anxiety symptoms and cortical amyloid deposition among cognitively normal elderly persons: The Mayo Clinic Study of Aging. Int Psychogeriatr 30, 245–251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Babulal GM, Roe CM, Stout SH, Rajasekar G, Wisch JK, Benzinger TLS, Morris JC, Ances BM (2020) Depression is associated with tau and not amyloid positron emission tomography in cognitively normal adults. J Alzheimers Dis 74, 1045–1055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Donovan NJ, Hsu DC, Dagley AS, Schultz AP, Amariglio RE, Mormino EC, Okereke OI, Rentz DM, Johnson KA, Sperling RA, Marshall GA (2015) Depressive symptoms and biomarkers of Alzheimer’s disease in cognitively normal older adults. J Alzheimers Dis 46, 63–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Gatchel JR, Donovan NJ, Locascio JJ, Schultz AP, Becker JA, Chhatwal J, Papp KV, Amariglio RE, Rentz DM, Blacker D, Sperling RA, Johnson KA, Marshall GA (2017) Depressive symptoms and tau accumulation in the inferior temporal lobe and entorhinal cortex in cognitively normal older adults: A pilot study. J Alzheimers Dis 59, 975–985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Babulal GM, Ghoshal N, Head D, Vernon EK, Holtzman DM, Benzinger TLS, Fagan AM, Morris JC, Roe CM (2016) Mood changes in cognitively normal older adults are linked to Alzheimer’s disease biomarker levels. Am J Geriatr Psychiatry 24, 1095–1104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Javaherian K, Newman BM, Weng H, Hassenstab J, Xiong C, Coble D, Fagan AM, Benzinger T, Morris JC (2019) Examining the complicated relationship between depressive symptoms and cognitive impairment in preclinical Alzheimer disease. Alzheimer Dis Assoc Disord 33, 15–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Gatchel JR, Rabin JS, Buckley RF, Locascio JJ, Quiroz YT, Yang H-S, Vannini P, Amariglio RE, Rentz DM, Properzi M, Donovan NJ, Blacker D, Johnson KA, Sperling RA, Marshall GA, Harvard Aging Brain Study (2019) Longitudinal association of depression symptoms with cognition and cortical amyloid among community-dwelling older adults. JAMA Netw Open 2, e198964–e198964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Chi S, Yu J-T, Tan M-S, Tan L (2014) Depression in Alzheimer’s disease: Epidemiology, mechanisms, and management. J Alzheimers Dis 42, 739–755. [DOI] [PubMed] [Google Scholar]
- [20].Chan CK, Soldan A, Pettigrew C, Wang J, Albert M, Rosenberg PB, BIOCARD Research Team (2020) Depressive symptoms and CSF Alzheimer’s disease biomarkers in relation to clinical symptom onset of mild cognitive impairment. Alzheimers Dement (Amst) 12, e12106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Chan CK, Soldan A, Pettigrew C, Wang M-C, Wang J, Albert MS, Rosenberg PB (2019) Depressive symptoms in relation to clinical symptom onset of mild cognitive impairment. Int Psychogeriatr 31, 561–569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Liew TM (2020) Neuropsychiatric symptoms in cognitively normal older persons, and the association with Alzheimer’s and non-Alzheimer’s dementia. Alzheimers Res Ther 12, 35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Hedden T, Mormino EC, Amariglio RE, Younger AP, Schultz AP, Becker JA, Buckner RL, Johnson KA, Sperling RA, Rentz DM (2012) Cognitive profile of amyloid burden and white matter hyperintensities in cognitively normal older adults. J Neurosci 32, 16233–16242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Vasquez BP, Zakzanis KK (2015) The neuropsychological profile of vascular cognitive impairment not demented: A meta-analysis. J Neuropsychol 9, 109–136. [DOI] [PubMed] [Google Scholar]
- [25].Albert M, Soldan A, Gottesman R, McKhann G, Sacktor N, Farrington L, Grega M, Turner R, Lu Y, Li S, Wang M-C, Selnes O (2014) Cognitive changes preceding clinical symptom onset of mild cognitive impairment and relationship to ApoE genotype. Curr Alzheimer Res 11, 773–784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].D’Ath P, Katona P, Mullan E, Evans S, Katona C (1994) Screening, detection and management of depression in elderly primary care attenders. I: The acceptability and performance of the 15 item Geriatric Depression Scale (GDS15) and the development of short versions. Fam Pract 11, 260–266. [DOI] [PubMed] [Google Scholar]
- [27].Kaufer DI, Cummings JL, Ketchel P, Smith V, MacMillan A, Shelley T, Lopez OL, DeKosky ST (2000) Validation of the NPI-Q, a brief clinical form of the Neuropsychiatric Inventory. J Neuropsychiatry Clin Neurosci 12, 233–239. [DOI] [PubMed] [Google Scholar]
- [28].Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC, Gamst A, Holtzman DM, Jagust WJ, Petersen RC, Snyde Peter J, Carrillo MC, Thies B, Phelps CH (2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR, Kawas CH, Klunk WE, Koroshetz WJ, Manly JJ, Mayeux R, Mohs RC, Morris JC, Rossor MN, Scheltens P, Carrillo MC, Thies B, Weintraub S, Phelps CH (2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Morris JC (1993) The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 43, 2412–2414. [DOI] [PubMed] [Google Scholar]
- [31].Soldan A, Pettigrew C, Fagan AM, Schindler SE, Moghekar A, Fowler C, Li Q-X, Collins SJ, Carlsson C, Asthana S, BIOCARD Research Team (2019) ATN profiles among cognitively normal individuals and longitudinal cognitive outcomes. Neurology 92, e1567–e1579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].DeCarli C, Maillard P, Fletcher E (2013) Four tissue segmentation in ADNI II. Department of Neurology and Center for Neuroscience, University of California, Davis. [Google Scholar]
- [33].Soldan A, Pettigrew C, Zhu Y, Wang M-C, Moghekar A, Gottesman RF, Singh B, Martinez O, Fletcher E, DeCarli C, Albert M, BIOCARD Research Team (2020) White matter hyperintensities and CSF Alzheimer disease biomarkers in preclinical Alzheimer disease. Neurology 94, e950–e960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Zhou Y, Resnick SM, Ye W, Fan H, Holt DP, Klunk WE, Mathis CA, Dannals R, Wong DF (2007) Using a reference tissue model with spatial constraint to quantify [11C]Pittsburgh compound B PET for early diagnosis of Alzheimer’s disease. Neuroimage 36, 298–312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Bilgel M, An Y, Zhou Y, Wong DF, Prince JL, Ferrucci L, Resnick SM (2016) Individual estimates of age at detectable amyloid onset for risk factor assessment. Alzheimers Dement 12, 373–379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Rothman KJ (1990) No adjustments are needed for multiple comparisons. Epidemiology 1, 43–46. [PubMed] [Google Scholar]
- [37].Nuño L, Gómez-Benito J, Carmona VR, Pino O (2021) A systematic review of executive function and information processing speed in major depression disorder. Brain Sci 11, 147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Steenland K, Karnes C, Seals R, Carnevale C, Hermida A, Levey A (2012) Late-life depression as a risk factor for mild cognitive impairment or Alzheimer’s disease in 30 US Alzheimer’s disease centers. J Alzheimers Dis 31, 265–275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Wilson RS, Arnold SE, Beck TL, Bienias JL, Bennett DA (2008) Change in depressive symptoms during the prodromal phase of Alzheimer disease. Arch Gen Psychiatry 65, 439–445. [DOI] [PubMed] [Google Scholar]
- [40].Sperling RA, Aisen PS, Beckett LA, Bennett DA, Craft S, Fagan AM, Iwatsubo T, Jack CR Jr, Kaye J, Montine TJ, Park DC, Reiman EM, Rowe CC, Siemers E, Stern Y, Yaffe K, Carrillo MC, Thies B, Morrison-Bogorad M, Wagster MV, Phelps CH (2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Barnes DE, Yaffe K, Byers AL, McCormick M, Schaefer C, Whitmer RA (2012) Midlife vs late-life depressive symptoms and risk of dementia: Differential effects for Alzheimer disease and vascular dementia. Arch Gen Psychiatry 69, 493–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Diniz BS, Butters MA, Albert SM, Dew MA, Reynolds CF (2013) Late-life depression and risk of vascular dementia and Alzheimer’s disease: Systematic review and meta-analysis of community-based cohort studies. Br JPsychiatry 202, 329–335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Misquitta K, Dadar M, Louis Collins D, Tartaglia MC, Alzheimer’s Disease Neuroimaging Initiative (2020) White matter hyperintensities and neuropsychiatric symptoms in mild cognitive impairment and Alzheimer’s disease. Neuroimage Clin 28, 102367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Anor CJ, O’Connor S, Saund A, Tang-Wai DF, Keren R, Tartaglia MC (2017) Neuropsychiatric symptoms in Alzheimer disease, vascular dementia, and mixed dementia. Neurodegener Dis 17, 127–134. [DOI] [PubMed] [Google Scholar]
- [45].Berlow YA, Wells WM, Ellison J, Sung YH, Renshaw PF, Harper DG (2010) Neuropsychiatric correlates of white matter hyperintensities in Alzheimer’s disease. Int J Geriatr Psychiatry 25, 780–788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [46].Kim HJ, Kang SJ, Kim C, Kim GH, Jeon S, Lee JM, Oh SJ, Kim JS, Choe YS, Lee KH, Noh Y, Cho H, Yoon CW, Chin J, Cummings JL, Lee JH, Na DL, Seo SW (2013) The effects of small vessel disease and amyloid burden on neuropsychiatric symptoms: A study among patients with subcortical vascular cognitive impairments. Neurobiology of Aging 34, 1913–1920. [DOI] [PubMed] [Google Scholar]
- [47].Hirono N, Kitagaki H, Kazui H, Hashimoto M, Mori E (2000) Impact of white matter changes on clinical manifestation of Alzheimer’s disease: A quantitative study. Stroke 31, 2182–2188. [DOI] [PubMed] [Google Scholar]
- [48].Mok VCT, Wong A, Wong K, Chu WCW, Xiong Y, Chan AYY, Kwok TCY, Hu X, Lee WK, Tang W, Wong KL, Wong S (2010) Executive dysfunction and left frontal white matter hyperintensities are correlated with neuropsychiatric symptoms in stroke patients with confluent white matter hyperintensities. Dement Geriatr Cogn Disord 30, 254–260. [DOI] [PubMed] [Google Scholar]
- [49].Insel PS, Weiner M, Mackin RS, Mormino E, Lim YY, Stomrud E, Palmqvist S, Masters CL, Maruff PT, Hansson O, Mattsson N (2019) Determining clinically meaningful decline in preclinical Alzheimer disease. Neurology 93, e322–e333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [50].Duron E, Hanon O (2008) Vascular risk factors, cognitve decline, and dementia. Vasc Health Risk Manag 4, 363–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [51].Tolppanen A-M, Solomon A, Soininen H, Kivipelto M (2012) Midlife vascular risk factors and Alzheimer’s disease: Evidence from epidemiological studies. J Alzheimers Dis 32, 531–540. [DOI] [PubMed] [Google Scholar]
- [52].Nelson PT, Dickson DW, Trojanowski JQ, Jack CR, Boyle PA, Arfanakis K, Rademakers R, Alafuzoff I, Attems J, Brayne C, Coyle-Gilchrist ITS, Chui HC, Fardo DW, Flanagan ME, Halliday G, Hokkanen SRK, Hunter S, Jicha GA, Katsumata Y, Kawas CH, Keene CD, Kovacs GG, Kukull WA, Levey AI, Makkinejad N, Montine TJ, Murayama S, Murray ME, Nag S, Rissman RA, Seeley WW, Sperling RA, White Iii CL, Yu L, Schneider JA (2019) Limbicpredominant age-related TDP-43 encephalopathy (LATE): Consensus working group report. Brain 142, 1503–1527. [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.