Harvard Aging Brain Study Research Articles

Cortical microstructure is associated with tau burden and predicts cognitive decline and clinical progression in healthy older adults

AbstractBackgroundNon‐invasive biomarkers of neuronal injury may be useful to detect cognitively‐normal individuals at risk of developing AD. A recent diffusion‐weighted imaging (DWI) method has allowed to assess microstructural properties of the brain's grey matter by means of cortical mean diffusivity (cMD) (Montal et al. Alz&Dem 14(3):340, 2018). Increased cMD is thought to reflect early neuronal injury that may precede macrostructural neurodegeneration. Here, we aimed to investigate the association of cMD with in vivo amyloid‐β and tau pathology in cognitively‐normal individuals, and to determine whether cMD can predict cognitive decline and clinical progression.Method196 cognitively‐normal participants from the Harvard Aging Brain Study (72.5±9.4 years; Table 1) underwent concurrent DWI, T1w‐MRI, 11C‐PIB‐PET, 18F‐flortaucipir (FTP)‐PET imaging and cognitive assessments at baseline. PIB‐PET was quantified using non‐PVC‐corrected Logan‐DVR, and FTP‐PET using PVC‐corrected SUVr; PET data were normalized to cerebellar grey‐matter. Longitudinal measures of Preclinical Alzheimer Cognitive Composite (PACC5) were available for n=186 individuals over a 3.72±1.96 yr follow‐up interval. Surface‐based image analysis (FreeSurfer 6.0) was performed to investigate vertex‐wise relationships between cMD and global amyloid‐β (PIB‐FLR), as well as tau (entorhinal, inferior‐temporal FTP). We assessed whether regional cMD predicted longitudinal PACC5 using linear mixed‐effects models, and progression to MCI using Cox proportional‐hazards regression.ResultEntorhinal and inferior‐temporal tau, but not amyloid‐β, were associated with increases in surface‐based cMD in AD‐signature clusters (Fig. 1A); illustrated for middle‐temporal gyrus cMD (Fig. 1B). Correcting for demographics and baseline PACC5, increased cMD within AD‐signature regions‐of‐interest predicted the rate of subsequent cognitive decline, which remained significant after correcting for PIB‐FLR and regional cortical thickness; there was a synergistic interaction between cMD and PIB‐FLR on the slope of PACC5 decline (Fig. 2). From n=163 participants with clinical follow‐up data, 11/163 (6.7%) converted to MCI within 3.3±1.5 years. Dichotomized regional cMD (top‐tertile group being "high cMD") predicted conversion to MCI, illustrated for the middle‐temporal gyrus, hazard ratio HR[95% CI] = 5.21 [1.14‐23.75], p=0.03, corrected by age, sex and education (Fig. 3).ConclusioncMD is a promising technique to quantify cortical microstructural changes possibly related to tau‐induced neuronal injury. Its ability to predict short‐term cognitive decline and clinical progression suggests utility in clinical trials.

Associations between biomarker status (amyloid, tau) and risk for progression to MCI/Dementia in the Harvard Aging Brain Study

AbstractBackgroundPrevious reports have shown that elevated AD biomarkers among clinically normal (CN) older adults confer heightened risk for a subsequent diagnosis of cognitive impairment using CSF and amyloid imaging. Here, we examine the impact of biomarker status on neuroimaging for both amyloid(A+/A‐) and tau(T+/T‐) on risk for clinical progression.MethodParticipants were CN at enrollment and underwent PET with 11CPiB to characterize global amyloid‐β(A) burden and F18Flortaucipir (FTP) to quantify inferior temporal tau(T). Participants were dichotomized as A+/A‐ and T+/T‐ using previously described mixture model distribution cut‐points. Because FTP only became available mid‐study, we undertook two analytical approaches: 1) cox regression analysis with first PiB scan as time origin (n=342, follow‐up=6.10 years;range=1‐9) with tau treated as a time‐varying covariate (setting it to T‐ prior to the first FTP scan if T‐ and setting it to missing prior to first scan if T+) and 2) first FTP scan as time origin with tau treated as fixed at that time (n=242, follow‐up=5.12 years;range=1‐7). Participants were diagnosed annually by biomarker‐blinded case conference. Covariates included Demographics and E4(+/‐).ResultIn the larger sample (31 individuals diagnosed with MCI/dementia), A+ was associated with increased diagnostic risk (HR=12.43, 95%CI=4.89‐31.57). In a model that included T as time‐varying (but not A), T+ was associated with increased risk for diagnosis (HR=4.46, 95% CI=0.99,20.11, p=0.051). Including A in the model, A+ was associated with increased diagnostic risk (HR=5.30, 95%CI=1.29,21.72) and T+ additionally appeared to confer higher risk, though not statistically significant (HR=2.89, 95%CI=0.12,67.73). An interaction term was not significant (HR for A+T+ versus A+T‐=4.65,95%CI=0.28,77.79). In the smaller sample with tau (19 individuals diagnosed), T+ was also associated with greater risk for diagnosis (HR=3.61, 95%CI=1.38‐9.40). Older age (HR=1.09, 95%CI=1.02‐1.18) and E4+ (HR=5.40, 95%CI=1.96‐14.89) were contributory. Both A+ (HR=13.83, 95%CI=2.82‐67.81) and T+ (HR=2.75, 95%CI=1.05‐7.24) are associated with higher diagnostic risk when included in the same model. A+T+ was associated with higher risk(HR=2.59, 95%CI=0.98‐6.83,p=0.055) compared with A+T‐.ConclusionA+ in combination with T+, is a significant risk factor for clinical progression. These findings provide further encouragement that interrupting A among CN older adults may mitigate risk for clinical progression.

Resilience and perceived stress in cognitively normal older adults during the COVID‐19 pandemic

BackgroundPsychological resilience, a construct representing a person’s ability to return to their baseline when difficulties occur, can play an important role in determining outcomes in the face of stressors. Resilience is a crucial component to consider during times of prolonged stress, such as experienced during the coronavirus disease (COVID‐19) pandemic, given that chronic stress has been associated with cognitive decline and dementia. Here, we investigated whether psychological resilience modified levels of perceived stress (assessed over 7 months), as well as examined the influence of demographic variables and Alzheimer’s disease biomarkers on this relationship.MethodIn May‐November 2020 of the COVID‐19 pandemic, 39 cognitively normal participants (mean age=78.38; 53.85% female) in the Harvard Aging Brain Study (HABS) completed questionnaires assessing stress (Perceived Stress Scale, PSS) and resilience (only at baseline; Connor‐Davidson Resilience Scale), higher scores on each measure indicating greater stress and increased resilience, respectively. We used previously collected data to better characterize the cohort, with the Preclinical Alzheimer Cognitive Composite (PACC;mean=0.46, SD=0.79) to assess cognition, and PET imaging data to assess Aß burden. A hierarchical regression model was used to assess whether resilience could predict baseline perceived stress. A separate linear mixed effect model (LME) was used to explore how resilience, time, and their interaction may affect stress. Both models also included age, sex, education, cognition, and Aß status (SUVR>1.324).ResultPerceived stress levels were generally low (mean=13.54, SD=8.65, range=0‐35), and did not significantly vary over the study period (p>0.5; mean number of timepoints=3.13). Both models demonstrated a main effect of resilience, such that higher resilience was associated with lower stress levels (p<0.001;Figure 1). Perceived stress was not influenced by demographic factors, cognition or Aß status. No significant interaction was found between resilience and time in predicting stress levels longitudinally.ConclusionOur findings confirm that resilience is critical for coping with stress during the COVID‐19 pandemic, suggesting that resilience could be an important target for potential interventions in the future. Efforts at increasing resilience could not only help support older adults navigating stressful situations, but may also mitigate negative effects of stress on neurocognitive and mental health.

Longitudinal trajectories of remote assessment of self‐ and study partner‐rated cognitive concerns, mood and Alzheimer’s disease biomarkers

AbstractBackgroundPrevious work has shown discordance between participant‐ and study‐partner‐reported cognitive concerns across symptomatic stages of Alzheimer’s Disease (AD). Longitudinal trajectories of participant and study‐partner report in preclinical AD have not been fully elucidated. This study sought to capture longitudinal trajectories of participant‐ and study‐partner‐reported concerns along with potential modifying factors: participant amyloid‐biomarker burden and depressive symptomsMethodParticipants came from the Harvard Aging Brain Study, a longitudinal observational cohort of cognitively unimpaired individuals at baseline who received PIB‐PET imaging (high‐amyloid&gt;1.185DVR;all participants and study‐partners blinded to amyloid status). In this sub‐study (n=74,mean age=77.29,56% female), every 3 months participants and study‐partners completed a modified Cognitive Function Index (CFI) measuring current concerns (median sessions completed=6) and participants also completed the Geriatric Depression Scale (GDS). Additionally, participants and study‐partners were asked about any change in cognitive functioning over the last year during an in‐clinic visit (median months=10). Linear mixed effects models were used to investigate the interaction between in‐clinic CFI and amyloid burden on longitudinal remote‐CFI, for both study‐partners and participants. Remote‐CFI slopes for participant and study‐partner were extracted from ordinary least‐squares regression models and compared to remote GDS slopes (stratified by participant amyloid status) using linear regression models. All models controlled for age, sex, and education.ResultParticipant remote‐CFI did not change over time (slope=0.94,SE=0.66,p=0.15), unlike the modestly‐increasing study‐partner remote‐CFI (slope=1.52,SE=0.66,p=0.02). Participant remote‐CFI slope was not significantly related to in‐clinic CFI. A significant interaction was seen with participant amyloid burden in study‐partners, whereby increasing study‐partner remote‐CFI was associated with higher in‐clinic CFI for participants with greater amyloid burden (estimate=‐1.21(0.6),p=0.04). In participants with high‐amyloid, however, concurrent change was seen in participant‐rated remote‐CFI and GDS, but not with study‐partner‐rated remote‐CFI.ConclusionIncreasing concerns over months are associated with an overall perception of change over the last year from study‐partners of higher‐amyloid participants. Conversely, in high‐amyloid participants, increasing cognitive concerns were associated with increasing depressive symptoms suggesting greater alignment between cognitive complaints and mood symptoms among those with high‐amyloid. These data suggest that frequent assessment offers rapid insight into emerging divergent patterns of cognitive concerns between participant and study‐partner during preclinical AD.

Trajectories of self‐rated concerns in individuals developing cognitive decline

AbstractBackgroundSubjective perception of cognitive decline is identified as an important marker related to the risk of progression to Alzheimer's disease. However, loss of awareness has also been observed in predementia stages, potentially reducing the validity of the subjective experience. The aim of this study was to characterize the trajectories of complaints in cognitively normal (CN) individuals who later developed cognitive decline, as well as investigate differences in demographics, brain pathology and clinical progression between individuals aware and unaware of these changes.MethodWe included participants (CN at baseline) from the Harvard Aging Brain Study. Linear Mixed effect models were used to compute individual longitudinal changes of memory complaints (7 questions questionnaire) and cognition (Preclinical Alzheimer’s Cognitive Composite, PACC5). Participants who demonstrated a statistically significant PACC5 decline over time were divided into two groups: an Unaware (N=36) group defined as individuals with statistically significantly decreasing or displaying low (&lt;2.5/7) complaints, and a Heightened (N=14) group defined as individuals with significantly increasing or displaying high (&gt;2.5/7) complaints (Fig 1). Participants who did not decline in memory were assigned to the Stable group (N=248). Kaplan‐Meier survival analysis was performed to determine the rate of progression from CN to clinically impaired (MCI or dementia diagnosis) across groups.ResultNo group differences were found for gender, education, depressive symptoms or Apolipoprotein E4‐carrier status (Tab 1). Unaware participants were significantly older than the Stable group (p&lt;0.001). The Heightened and Unaware groups demonstrated equivalent amyloid burden, which was significantly greater than the Stable group (p&lt;0.01). The survival analysis revealed significant differences between all groups (p&lt;0.001,Fig 2). Heightened participants had the fastest rate of clinical progression, followed by the Unaware group.ConclusionParticipants with heightened awareness demonstrated the fastest rate of progression from cognitively normal to MCI or dementia. Importantly, among the individuals with a subsequent cognitive decline, a number of individuals (72%) were unaware of their decline. These individuals had the same demographic profile and amyloid burden as the individuals with heightened awareness. Future research will investigate the mechanisms of self‐awareness and potential dysfunction of systems underlying the loss of awareness, as it can be essential for the preservation of individual autonomy.

Regional beta‐amyloid and tau deposition: Results from the Framingham Heart Study

AbstractBackgroundThe Framingham Heart Study (FHS), a three‐generation community‐based cohort studying cardiovascular disease across the adult lifespan, has expanded to include positron emission tomography (PET) of beta‐amyloid (Aβ) and tau. Our aim is to characterize these pathologies in this unique sample and utilize the wide age range to assess the spatiotemporal ordering of emerging Aβ and tau.Method211 clinically‐normal adults aged 33‐74 from FHS underwent Pittsburgh Compound B (PIB) and Flortaucipir (FTP) PET. PIB and FTP were regionally quantified in Desikan regions using distribution volume ratio (DVR) for PIB and standardized uptake volume ratio (SUVR) for FTP. Three approaches were used to identify early accumulating regions: 1. A series of Age*APOE linear models to identify regions accumulating earlier in the lifespan in ε4 carriers. 2. Identification of early regions as those more frequently elevated as previously used in older adults, with Gaussian mixture models (GMM) conducted to evaluate bimodality and set biomarker positivity thresholds for each region. 3. Applied the same frequency approach using well‐validated PIB GMM thresholds based on older adults from the Harvard Aging Brain Study (HABS).ResultFor PIB, significant Age*APOE interactions (Fig. 1) were seen across multiple regions previously implicated as early‐accumulating based on elevation frequencies from older adult samples, as well as the pars opercularis (p=0.005) and lateral parietal cortices (p&lt;0.04). PIB was significantly bimodal across all ROIs (LRT&gt;36.1, p&lt;0.001; Fig. 1). Application of GMM‐based regional positivity frequencies derived within FHS versus from HABS thresholds led to substantial discrepancies in the spatiotemporal ordering of PIB (rs=0.30, p=0.17).Age*APOE effects were not detected for any tau regions, but inferior temporal (IT), middle temporal (MT), and amygdala (AM) increased with age (Fig. 2). Higher global Aβ was also associated with higher FTP SUVR in IT, MT, and a trend for AM, as well as inferior parietal and precuneus (Fig. 2).ConclusionAβ and tau mainly follow spatiotemporal patterns consistent with prior evidence in older adult samples. GMM‐based methods used in older adult samples to dichotomize biomarker positivity may be suboptimal in younger samples. Continuous approaches may better capture early Aβ and tau in younger populations.

Greater psychological resilience during the COVID‐19 pandemic is associated with lower tau burden in cognitively unimpaired individuals

BackgroundThe outbreak of the 2019 novel coronavirus SARS‐CoV‐2, causing the COVID‐19 pandemic, along with the global measures used to control the spread of the disease, have been particularly stressful for older individuals. Robust evidence suggests that stress‐related physiological processes may play a significant role in the onset of age‐related cognitive decline and dementia, such as Alzheimer’s disease (AD). For example, previous animal and human research have shown that stress exacerbates tau pathology and subsequent cognitive impairment. One important factor determining a person’s level of stress is psychological resilience, which refers to a person’s ability to return to equilibrium when difficulties occur. As such, resilience may buffer and protect individuals against the deleterious effects of stress. The objectives of this study were to examine the relationship between psychological resilience and tau burden in cognitively unimpaired individuals, and to evaluate whether amyloid (Aß) pathology modifies this relationship.MethodA total of 114 older adults (mean age=74.5;66 females (57.9%)) enrolled in the Harvard Aging Brain Study or the Instrumental Activities of Daily Living study completed an electronic survey online including measures of psychological resilience (Connor‐Davidson Resilience Scale‐10, where a higher score indicates greater resilience) during the COVID‐19 pandemic period (May 2020). We used previously collected (mean time from survey=1.75 year) PET imaging data to assess Aß and tau burden.ResultsOverall, participants demonstrated high levels of resilience (mean=29.5,range=9‐40). Multiple regression analyses (correcting for age, gender and time difference between survey and PET) demonstrated an interaction between resilience and amyloid on tau pathology (ß=‐0.01,p=0.015, Figure1). Specifically, in Aß+ individuals (SUVR>1.324), those with higher psychological resilience also had lower tau pathology. No relationship was seen in Aß‐ individuals.ConclusionOur findings suggest that greater resilience during the pandemic is associated with lower tau burden in individuals at higher risk of cognitive decline (i.e. preclinical AD). Future studies are needed to determine the impact of stress and psychological resilience on prospective pathological brain changes and subsequent cognitive decline. Furthermore, resilience may be an important target for interventions to support older adults navigating stressful situations and to lessen the influence of stressors on AD pathological processes.

Extraneous neuroimaging factors do not contribute to sex differences in flortaucipir signal: Analysis of skull binding and partial volume effects

AbstractBackgroundClinically normal females exhibit greater [18]F‐flortaucipir (FTP) PET signal than males in both temporal and neocortices. It remains unclear whether sex differences in neocortical regions are primarily explained by technical variability issues. We aimed to investigate the contribution of signal spillover/off‐target skull binding to sex differences in FTP‐PET. Next, we explored partial volume effects (PVE) by simulating sex differences in smoothed FTP‐PET signal. Discerning sex differences in tau signal versus noise is pivotal to understanding sex differences in the pathology of Alzheimer’s disease and associated tauopathies.Method343 clinically normal (female=58%; mean[SD]=73.8[8.5] years) (female=38%; mean[SD]=76.9[7.3] years) participants from the Harvard Aging Brain Study and the Alzheimer’s Disease Neuroimaging Initiative underwent cross‐sectional FTP‐PET (standardized uptake value ratios [SUVrs]). For skull analyses, we created skull ROIs based on signal 12mm from the outer perimeter of voxels in FreeSurfer‐defined tau ROIs. Linear regression models estimated the main effects of sex across cortical tau ROIs while correcting for local skull binding. We simulated PVE by convolving group‐level SUVr means in each ROI with 6mm Gaussian kernels, and then compared the sexes with linear regression models post‐smoothing.ResultWidespread sex differences in skull binding were observed (Table 1). Covarying for skull binding ameliorated weaker sex differences in cortical FTP signal but did not impact the largest effects. Sex differences in PVE were observed in both female and male directions; no clear sex‐related biases in PVE were found to impact cortical tau sex differences except for the rostral middle frontal region (Figure 1).ConclusionOur findings suggest that sex differences in FTP‐PET are not solely attributed to skull ‘clouding’ or PVE, but rather support hypotheses of female‐related tau vulnerability. Nevertheless, as only two potential confounds were investigated, and gross morphology/volumetric issues remain a key concern, further investigation is needed to fully elucidate this phenomenon. Investigations of sex differences in longitudinal tau accumulation (a few preliminary reports already suggest faster rates in females) will add further support to the argument that noise properties inherent in FTP‐PET do not significantly contribute to sex differences in cortical tau signal.

Monthly computerized at‐home assessments to detect cognitive change in preclinical Alzheimer’s disease

AbstractBackgroundAlongside the increased focus on detecting Alzheimer’s disease (AD) pathology in the preclinical stage, there is a need to more rapidly track AD‐related cognitive changes that may emerge during this stage. Computerized cognitive testing has the potential to achieve this by enabling more frequent, remote, unsupervised assessments. Here, we aimed to investigate whether monthly at‐home assessments of a modified version of the Behavioural Pattern Separation Task‐Object Version (BPSO) could detect cognitive change in cognitively unimpaired (CU) individuals with and without abnormal AD biomarkers.MethodN=110 CU participants (age=77.1±4.9, 61% female, MMSE 29±1.3) from the Harvard Aging Brain Study completed the BPSO monthly at‐home on an iPad as part of the Cogstate C3 battery for up to one year (6.3±3.9 months follow‐up). All participants underwent PIB‐PET imaging (1.25±1.05 years before at‐home baseline) and a subset (n=82) underwent Flortaucipir PET (0.56±0.4 years before at‐home baseline). Primary outcome of the BPSO is a metric reflecting the ability to correctly discriminate between stimuli that are similar but not identical to previously learned targets (score range 0‐1, higher scores reflecting better performance). Linear mixed models were used to characterize BPSO performance over months (including both linear and quadratic time terms to investigate nonlinear change), and to examine associations with amyloid (dichotomous) and medial‐temporal tau deposition (continuous) while adjusting for age, sex, education and their interactions with time.ResultOverall, individuals’ BPSO performance improved over months, as shown by a significant quadratic effect of time (β=‐.002,p&lt;.001) with a plateauing learning curve (Figure 1). On average, amyloid‐positive individuals (n=29) showed slightly worse BPSO performance (β=‐.06,p=.03) than amyloid‐negative individuals, but similar trajectories over time. However, the baseline amyloid effect was attenuated when limiting the sample to those with both amyloid and tau imaging. Less improvement over the year was associated with greater tau burden in the entorhinal cortex (Time*tau β=‐0.02,p=.019) (Figure 2).ConclusionWe show that subtle alterations in pattern separation performance over repeated exposure in CU individuals are associated with tau deposition in the medial‐temporal lobe. This implies that unsupervised computerized testing using monthly learning paradigms may be an efficient way to early detect cognitive change associated with preclinical AD.

Association of cortical microstructure with amyloid-\u03b2 and tau: impact on cognitive decline, neurodegeneration, and clinical progression in older adults

Noninvasive biomarkers of early neuronal injury may help identify cognitively normal individuals at risk of developing Alzheimer’s disease (AD). A recent diffusion-weighted imaging (DWI) method allows assessing cortical microstructure via cortical mean diffusivity (cMD), suggested to be more sensitive than macrostructural neurodegeneration. Here, we aimed to investigate the association of cMD with amyloid-β and tau pathology in older adults, and whether cMD predicts longitudinal cognitive decline, neurodegeneration and clinical progression. The study sample comprised n = 196 cognitively normal older adults (mean[SD] 72.5 [9.4] years; 114 women [58.2%]) from the Harvard Aging Brain Study. At baseline, all participants underwent structural MRI, DWI, 11C-Pittsburgh compound-B-PET, 18F-flortaucipir-PET imaging, and cognitive assessments. Longitudinal measures of Preclinical Alzheimer Cognitive Composite-5 were available for n = 186 individuals over 3.72 (1.96)-year follow-up. Prospective clinical follow-up was available for n = 163 individuals over 3.2 (1.7) years. Surface-based image analysis assessed vertex-wise relationships between cMD, global amyloid-β, and entorhinal and inferior-temporal tau. Multivariable regression, mixed effects models and Cox proportional hazards regression assessed longitudinal cognition, brain structural changes and clinical progression. Tau, but not amyloid-β, was positively associated with cMD in AD-vulnerable regions. Correcting for baseline demographics and cognition, increased cMD predicted steeper cognitive decline, which remained significant after correcting for amyloid-β, thickness, and entorhinal tau; there was a synergistic interaction between cMD and both amyloid-β and tau on cognitive slope. Regional cMD predicted hippocampal atrophy rate, independently from amyloid-β, tau, and thickness. Elevated cMD predicted progression to mild cognitive impairment. Cortical microstructure is a noninvasive biomarker that independently predicts subsequent cognitive decline, neurodegeneration and clinical progression, suggesting utility in clinical trials.

Comparing PET and MRI Biomarkers Predicting Cognitive Decline in Preclinical Alzheimer Disease.

To compare how structural MRI, fluorodeoxyglucose (FDG), and flortaucipir (FTP) PET signals predict cognitive decline in high-amyloid vs low-amyloid participants with the goal of determining which biomarker combination would result in the highest increase of statistical power for prevention trials. In this prospective cohort study, we analyzed data from clinically normal adults from the Harvard Aging Brain Study with MRI, FDG, FTP, and Pittsburgh compound B (PiB)-PET acquired within a year and prospective cognitive evaluations over a mean 3-year follow-up. We focused analyses on predefined regions of interest: inferior temporal, isthmus cingulate, hippocampus, and entorhinal cortex. Cognition was assessed with the Preclinical Alzheimer's Cognitive Composite. We evaluated the association between biomarkers and cognitive decline using linear mixed-effect models with random intercepts and slopes, adjusting for demographics. We generated power curves simulating prevention trials. Data from 131 participants (52 women, age 73.98 ± 8.29 years) were analyzed in the study. In separate models, most biomarkers had a closer association with cognitive decline in the high-PiB compared to the low-PiB participants. A backward stepwise regression including all biomarkers demonstrated that only neocortical PiB, entorhinal FTP, and entorhinal FDG were independent predictors of subsequent cognitive decline. Power analyses revealed that using both high PiB and low entorhinal FDG as inclusion criteria reduced 3-fold the number of participants needed in a hypothetical trial compared to using only high PiB. In preclinical Alzheimer disease, entorhinal hypometabolism is a strong and independent predictor of subsequent cognitive decline, making FDG a potentially useful biomarker to increase power in clinical trials. This study provides Class II evidence that in people with preclinical Alzheimer disease, entorhinal hypometabolism identified by FDG-PET is predictive of subsequent cognitive decline.

Longitudinal predictive modeling of tau progression along the structural connectome

Tau neurofibrillary tangles, a pathophysiological hallmark of Alzheimer’s disease (AD), exhibit a stereotypical spatiotemporal trajectory that is strongly correlated with disease progression and cognitive decline. Personalized prediction of tau progression is, therefore, vital for the early diagnosis and prognosis of AD. Evidence from both animal and human studies is suggestive of tau transmission along the brains preexisting neural connectivity conduits. We present here an analytic graph diffusion framework for individualized predictive modeling of tau progression along the structural connectome. To account for physiological processes that lead to active generation and clearance of tau alongside passive diffusion, our model uses an inhomogenous graph diffusion equation with a source term and provides closed-form solutions to this equation for linear and exponential source functionals. Longitudinal imaging data from two cohorts, the Harvard Aging Brain Study (HABS) and the Alzheimer’s Disease Neuroimaging Initiative (ADNI), were used to validate the model. The clinical data used for developing and validating the model include regional tau measures extracted from longitudinal positron emission tomography (PET) scans based on the 18F-Flortaucipir radiotracer and individual structural connectivity maps computed from diffusion tensor imaging (DTI) by means of tractography and streamline counting. Two-timepoint tau PET scans were used to assess the goodness of model fit. Three-timepoint tau PET scans were used to assess predictive accuracy via comparison of predicted and observed tau measures at the third timepoint. Our results show high consistency between predicted and observed tau and differential tau from region-based analysis. While the prognostic value of this approach needs to be validated in a larger cohort, our preliminary results suggest that our longitudinal predictive model, which offers an in vivo macroscopic perspective on tau progression in the brain, is potentially promising as a personalizable predictive framework for AD.

Early Detection and Treatment of Age-Related Cognitive Decline

Jennifer Gatchel (Symposium Co-chair) will introduce the speakers and then present data from the Harvard Aging Brain Study and screening data from the Anti-Amyloid Treatment in Asymptomatic Alzheimer's Disease (A4) Study. Mood, cognition, and PET data from 4,486 participants will be presented from the A4 Study and from 135 participants from the Harvard Aging Brain Study. These unpublished data show associations among neuropsychiatric symptoms, cognition, and in vivo neuroimaging biomarkers—amyloid and tau—in cognitively unimpaired individuals. Findings support cortical amyloid and regional cerebral tau as risk factors for cognitive decline in older adults. Davangere Devanand will review the evidence that olfactory identification deficits predict the development of Alzheimer's dementia, and show that olfactory impairments are superior to episodic verbal memory deficits in predicting cognitive decline. He will present data from a large community sample showing that intact performance on brief global cognitive and olfactory tests is associated with lack of transition to dementia during follow-up. These unique data will show that in 711 subjects, intact olfactory and cognitive abilities were protective with few transitions to dementia and none in the 70-75 and 81-83 years age groups. Terry Goldberg will review the apolipoprotein E e4 allele (APOE e4) as a risk factor for Alzheimer's dementia, and present unpublished data from the National Alzheimer's Consortium database showing that the apolipoprotein E e2 (APOE e2) genotype is protective against amyloid and tau pathologies of Alzheimer's disease but not Lewy body pathology, fronto-temporal or related tauopathy dementias as identified by gold standard neuropathological diagnosis. He will describe unpublished genetic and pathologic results from the Alzheimer's NACC consortium (N=1557) showing that APOE e2 protects against Alzheimer's and APOE e4 promotes Lewy body pathology in limbic and cortical regions but not midbrain. Gary Small (Symposium Chair) will moderate the audience questions and panel responses after presenting new data from randomized controlled trials indicating the cognitive benefits of pomegranate juice, bioavailable curcumin, or physical exercise – all strategies that reduce inflammation. These findings, as well as previous research on medications, physical activity, diet, obesity, and other risk/protective factors, suggest that several anti-inflammatory interventions may protect brain health and lower the risk for Alzheimer's dementia. He will describe data from placebo-controlled trials of memory effects of pomegranate juice (N=260) and curcumin (N=40), and a trial of memory training/aerobic exercise (N=55).

Association of Digital Clock Drawing With PET Amyloid and Tau Pathology in Normal Older Adults.

To determine whether a digital clock-drawing test, DCTclock, improves upon standard cognitive assessments for discriminating diagnostic groups and for detecting biomarker evidence of amyloid and tau pathology in clinically normal older adults (CN). Participants from the Harvard Aging Brain Study and the PET laboratory at Massachusetts General Hospital were recruited to undergo the DCTclock, standard neuropsychological assessments including the Preclinical Alzheimer Cognitive Composite (PACC), and amyloid/tau PET imaging. Receiver operating curve analyses were used to assess diagnostic and biomarker discriminability. Logistic regression and partial correlations were used to assess DCTclock performance in relation to PACC and PET biomarkers. A total of 300 participants were studied. Among the 264 CN participants, 143 had amyloid and tau PET imaging (Clinical Dementia Rating [CDR] 0, Mini-Mental State Examination [MMSE] 28.9 ± 1.2). An additional 36 participants with a diagnosis of mild cognitive impairment or early Alzheimer dementia (CDR 0.5, MMSE 25.2 ± 3.9) were added to assess diagnostic discriminability. DCTclock showed excellent discrimination between diagnostic groups (area under the receiver operating characteristic curve 0.86). Among CN participants with biomarkers, the DCTclock summary score and spatial reasoning subscores were associated with greater amyloid and tau burden and showed better discrimination (Cohen d = 0.76) between Aβ± groups than the PACC (d = 0.30). DCTclock discriminates between diagnostic groups and improves upon traditional cognitive tests for detecting biomarkers of amyloid and tau pathology in CN older adults. The validation of such digitized measures has the potential of providing an efficient tool for detecting early cognitive changes along the AD trajectory. This study provides Class II evidence that DCTclock results were associated with amyloid and tau burden in CN older adults.

Neuroimaging correlates of Stages of Objective Memory Impairment (SOMI) system.

IntroductionTo assess the relationship between memory performance defined by the Stages of Objective Memory Impairment (SOMI) system and the Alzheimer's disease (AD) ATN (amyloid beta [A], pathologic tau [T], and neurodegeneration [N]) biomarker system.MethodsWe used data from the Harvard Aging Brain Study cohort to estimate the level of ATN biomarkers: amyloid beta (C‐Pittsburgh compound B‐positron emission tomography [PET]), tau (F‐18–flortaucipir [FTP] PET), and neurodegeneration (magnetic resonance imaging volumetrics). We assessed the cross‐sectional relationship of SOMI classification with global amyloid levels, entorhinal and inferior temporal tau deposition, and hippocampal atrophy.ResultsParticipants with both memory storage and retrieval deficits (SOMI‐3, ‐4) had smaller hippocampal volumes and higher entorhinal and inferior temporal tau burden than participants with no memory impairment (SOMI‐0) or mild retrieval difficulty (SOMI‐1). Amyloid burden did not differ among SOMI stages.DiscussionThis pilot supports the close relationship between tau pathology and memory impairment across the AD continuum. SOMI may be useful to determine eligibility for randomized controlled trials prior to the assessment of biomarker status.

The dynamic interplay between longitudinal subjective and objective cognitive decline along the early AD spectrum in the Harvard Aging Brain Study

AbstractBackgroundApplying a biomarker‐based definition of Alzheimer’s disease (AD) (Jack et al., 2018) affords characterization of early, disease‐specific symptoms that are subtle, yet dynamic. Our previous work comparing cognitively‐unimpaired individuals with elevated amyloid (Ab+) to non‐elevated (Ab‐), shows increasing subjective cognitive decline, while neuropsychological performance decreases longitudinally. Here, we sought to directly examine longitudinal associations between subjective and objective cognitive decline along the early AD trajectory.Method224 cognitively‐unimpaired individuals at baseline from the Harvard Aging Brain Study, a community‐based sample, were followed up to 9 years with annual neuropsychological assessment (PACC) and a participant reported cognitive complaint questionnaire (STIDA). A second cognitive complaint questionnaire (CFI) was introduced several years into the study (mean=3.23 (1.65) years) and was administered to both the participant and a study partner annually thereafter. Ab‐PET using Pittsburgh Compound B was collected closest to the time of CFI. Previously established cut‐offs were used to determine Ab+/‐. Longitudinal trajectories of STIDA, PACC, and CFI (participant and study partner) were examined by Ab status.ResultPiecewise linear mixed models revealed initially increasing STIDA (0‐3.5years, t=2.17, p = 0.03) that plateaued in the remaining years (3.6‐9.1 years, t = 0.26, p = 0.80) in Ab+. Likewise, increasing STIDA corresponded with decreasing PACC in the first 3 years (r = ‐0.36, p = 0.001), but not in subsequent years (r = 0.11, p = 0.33). When examining cross‐sectional CFI (first collection of CFI) with retrospective PACC (mean= 3.23 years), higher participant CFI (r = ‐0.26, p = 0.03) and study partner CFI (r = ‐0.32, p = 0.006) were associated with steeper PACC decline in Ab+. Finally, comparison of concurrent slopes between longitudinal participant and study partner CFI, revealed a significant divergent pattern (t=‐2.50, p= 0.01), such that study partner CFI increases relative to slightly decreasing participant CFI over time.ConclusionAmong Ab+ individuals, participant report initially increases and tracks with objective decline. Over time, however, participant report begins to diminish as study partner report continues to increase. Our findings highlight the dynamic nature of subjective report in early AD, which has significant implications for clinically meaningful outcome measures in secondary prevention trials.

Harmonizing the preclinical Alzheimer cognitive composite for multi‐cohort studies

AbstractBackgroundVersions of the Preclinical Alzheimer Cognitive Composite (PACC) and within cohort standardization models have been implemented in multiple observational and clinical trials. Harmonizing PACC data across studies, however, requires modern psychometric approaches to reflect the psychometric, testing, and demographic nuances between cohorts and enable meta‐analyses. We present a method for harmonizing PACC across four cohorts: the Alzheimer’s Disease Neuroimaging Initiative (ADNI), the Anti‐Amyloid Treatment in Asymptomatic Alzheimer’s disease (A4), Harvard Aging Brain Study (HABS), and Australian Imaging, Biomarker &amp; Lifestyle Study of Ageing (AIBL).MethodTables 1 and 2 show the demographic and neuropsychological test data used to form each PACC. The ADNI dataset was defined as the anchor for harmonization. The final visit data were used to set factor score loadings for each PACC (except A4) to maximize cognitive variability in the sample for calibration purposes. First, we used confirmatory factor analysis (CFA) on the ADNI dataset to ascertain raw loadings and thresholds of each test onto a latent PACC factor (lPACC). These loadings were then applied to the longitudinal ADNI dataset (Fig.1). For the other studies, common items (MMSE and Logical Memory) were set as “anchors” to freely estimate unique items. Common items in HABS were set as anchors for A4 (FCsrt and DSST). Longitudinal lPACC scores were generated once all item parameters were estimated for all cohorts. We performed validation analyses to assess the quality of integration using lPACC scores versus within‐cohort standardized PACC (zPACC).ResultLPACC scores largely aligned with zPACC, particularly when examining within‐cohort relationships with baseline demographic variables (Table3) and comparing slope distributions across cohorts (Fig2B). Notable differences between the scores include: a) shifted baseline distributions for lPACC relative to zPACC, reflecting cohort demographic heterogeneity (Fig.2A), and b) constricted slope distribution for lPACC slopes compared with zPACC slopes (Fig.2B), suggesting reduced variability (Fig.2B/C). When the cohorts were combined, the point estimate for strength of association between baseline amyloid and longitudinal cognitive change was higher using the lPACC than the zPACC (Fig.3).ConclusionLPACC scores reflect within cohort patterns, while reducing slope variability. Using advanced psychometrics, a harmonized PACC can be pooled across studies, facilitating robust multi‐cohort analyses.

Repeated memory‐based assessments: Implications for clinical trials and practice

AbstractBackgroundDetailed longitudinal neuropsychological assessment in the Harvard Aging Brain Study (HABS) and related studies has revealed that subtle decrements in associative and semantic memory, as well as reduced practice effects for repeated items may be the earliest cognitive changes in preclinical AD. Utilizing digital, repeated, and remote assessment of these cognitive functions has the potential to increase the rapidity and ease with which these subtle changes can be detected. We previously showed that diminished practice effects on a challenging iPad‐based monthly face‐name memory paradigm was associated with neuroimaging markers of amyloid and tau in normal older adults participating in HABS. These findings inspired the development of the Boston Remote Assessment of Neurocognitive Health (BRANCH), a smartphone‐based associative and semantic memory assessment using stimuli and tasks relevant to everyday life designed for serial assessment.MethodWe describe the development, feasibility, and pilot data from a smartphone‐based memory assessment administered daily (over consecutive days) in normal older adults using their own devices. BRANCH includes 2 measures of paired associative learning (faces and names, groceries and prices), an associative memory test with facilitated encoding (categories), and a continuous visual recognition task (street signs). A total of 41 pilot participants (mean age=76.1; 64% female) completed 1 version of BRANCH in clinic and 4 versions remotely on their own mobile device for 4 consecutive days. Participants also completed standard paper and pencil testing and a questionnaire about their experience completing the task.ResultBRANCH was feasible with a total of 80% of individuals completing all assessments in the correct order. It was acceptable to participants with 64% finding the tasks at‐least somewhat to highly engaging. Older age and general cognition (e.g., MMSE) was associated with worse performance across memory tasks. Improved performance (practice effect) was observed on daily administration of paired associative memory tasks (Figure 1.).ConclusionA theoretically and evidence‐based digital memory assessment with ecologically‐valid tasks and stimuli is feasible for older adults to complete independently on their own devices. Digital assessment over multiple timepoints has the potential to drastically improve the efficiency with which cognitive performance and subtle decline can be captured.

Faster rates of tau accumulation in FTP‐PET in females relative to males, and a cross‐sectional influence on faster cognitive decline: Preliminary findings from HABS and ADNI

AbstractBackgroundClinically normal older females exhibit higher cross‐sectional 18F‐Flortaucipir (FTP)‐PET signal, but not Aβ‐PET, than males, particularly in temporal regions. What remains unclear is the extent to which these findings represent a real biological phenomenon as opposed to a methodological bias. Here, we explored sex differences in longitudinal tau accumulation, and how contemporaneous rates of FTP‐PET, Aβ‐PET and cognitive change may be modified by sex.MethodWe examined 128 clinically‐normal participants from the Harvard Aging Brain Study (HABS; 60% Female, Age=75(6.6)years) with at least two time‐points of contemporaneous FTP‐PET, PiB‐PET and cognition. Also, 153 ADNI participants with at least two time‐points of FTP‐PET (nCN=94/nMCI=59; 50% Female, Age=73(6.4)years), and with prospective longitudinal AV45‐PET and cognition (n=131). Mean follow up was 3.3yrs (range=1.2‐6.2yrs) and 1.4yrs (range=0.6‐3.3yrs), respectively. For FTP‐PET we exclusively examined the inferior temporal lobe (non‐partial volume corrected, cerebellar greyreference), for Aβ‐PET we used the PiB summary measure in HABS (cerebellar greyreference) and the AV‐45 summary measure in ADNI (whole cerebellumreference). For cognition, we used the Preclinical Alzheimer Cognitive Composite (PACC). We examined relationships between contemporaneous longitudinal FTP‐PET, Aβ‐PET, and cognitive performance using linear mixed models, adjusting for age, years of education and random intercept.ResultIn HABS, females exhibited faster rates of FTP‐PET accumulation than males, but only with increasing rates of Aβ‐PET accumulation (t=1.93, p=0.003; see Figure 1A). In ADNI, females exhibited faster rates of FTP‐PET accumulation, not moderated by Aβ‐PETchange (t=2.92, p=0.004; see Figure 1B). Sex did not moderate the relationship between FTP‐PET and PACC change in HABS or ADNI (see Figure 2.). We did, however, find that females with higher baseline FTP‐PET resulted in faster PACC decline than men in only clinically‐normal individuals in both cohorts (t=‐1.86ADNI,‐2.72HABS, p∼0.02; see Figure 3.).ConclusionThese findings support a female susceptibility to tau as a biological phenomenon of accelerated tau accumulation in AD‐vulnerable regions. A lack of an association between changing tau and changing cognition remains to be understood. It is possible that our models are picking up more on the ‘severity of burden’ (e.g. baseline tau) to influence faster rates of decline in females, rather than dynamic change per se.

Are amyloid and tau synergistic? How to interpret an amyloid/tau interaction on cognitive decline in clinically normal adults

AbstractBackgroundMany recent studies have examined interactions between β‐amyloid (Aβ) and tau‐PET to predict clinical progression. Ab and tau are moderately correlated in clinically‐normal older adults (r∼0.4‐0.5, p&lt;0.001; Figure 3B), which complicates statistical modelling of interactive effects. Multi‐collinearity can result in conflated interactions that reflect nonlinear effects (square of shared variance) rather than true interactions between separate channels of information. Here, we use simulated and real data to examine the extent to which an Aβ and tau interaction reflects ‘true’ synergism vs. conflated shared variance in predicting cognitive decline.MethodWe first ran simulations to illustrate how interactions between collinear variables inflates the association with shared sources of variance: we simulated two collinear independent variables by generating three random sources of variance (one shared variance [v1] and two unique sources of variance [v2 and v3]) (Figure 1). extracted two principal components from these simulated variables (PC1shared and PC2unique) and ran correlations between a PC1*PC2 interaction and different combinations of the simulated v1‐v3 sources of variance (Figure 1). We also examined real data (Aβ‐PET, entorhinal and inferior temporal tau‐PET, and longitudinal PACC) from the Harvard Aging Brain Study (n=224; Agemean=72.0 (sd=7.3), Female=59%) and ADNI (n=291; Agemean=71.8 (sd=6.5), Female=47%), in order to compare the pattern of effects with the simulated results.ResultThe PC1shared, and its square, PC1shared 2, were strongly associated with shared variance and its square (v1 and v1 2; see Figure 2). A Principal Components Analysis (PCA) of Aβ and entorhinal tau in HABS and ADNI cohorts resulted in a PC1shared and PC1shared 2 that closely associated with PACC slopes, mirroring the simulated results showing collinear interactions are more likely to represent conflated shared variance (see Figure 3A). Interactions between Aβ with inferior temporal tau were less extreme, suggesting additional associations between PACC slopes and PC2unique.ConclusionOur results suggest an Aβ and tau interaction should be interpreted with caution, as this may represent the non‐linear effects of shared variance, rather than the interaction of two (Aβ and tau) unique sources of variance. That is, PC1shared can be interpreted as a squared term, e.g. tau2 or Aβ2, rather than reflecting something truly synergistic between the proteinopathies to influence cognitive decline.