AbstractBackgroundThe U.S. Study to Protect Brain Health Through Lifestyle Intervention to Reduce Risk (U.S. POINTER) is a two‐year randomized controlled trial designed to evaluate the effects of lifestyle interventions on older adults aged 60‐79 years. The POINTER Imaging ancillary study employs arterial spin labeling (ASL) magnetic resonance imaging (MRI) with two inversion times, enabling the mapping of arterial transit time (ATT) and the investigation of its relationship with dementia risk factors. This study examines the effect of the Framingham Risk Score (FRS) on ATT, which provides hemodynamic information on the cerebral vasculature.MethodA subset of POINTER Imaging participants (n = 134; F/M = 81/53; age = 68.1±5.1; FRS = 11.3±10.4) from two imaging sites underwent two ASL MRI scans during the baseline session. Two 3D Pulsed ASL scans were obtained with TI1 of 790ms and TI2 values of 2000ms or 2750ms, providing delay times (between labeling and imaging) of 1210ms or 1960ms. ATT, the time required for blood to travel from the neck to the tissue, was estimated in the range of 1210ms and 1960ms using a kinetic model (Figure 1). ATT maps were normalized to a standard space, and a voxel‐wise linear regression (ATT∼age+sex+log(FRS)) was performed on ATT maps in gray matter to assess the effect of FRS on ATT. To control false positives, clusters smaller than 1.4ml were removed based on a simulation with α = 0.05.ResultFigure 2 depicts three clusters found in the occipital, precuneus and frontal gray matter regions, all of which demonstrate a positive relationship between ATT and log(FRS). Participants with higher FRS exhibit prolonged mean ATT values in the clusters as shown in Figure 3. After stratifying into sex groups, both groups show the same direction.ConclusionAlthough the detectable range of ATT with two ASL scans was limited, the analysis suggests that an increased risk of cardiovascular disease, as quantified by the Framingham Risk Score, can serve as an indicator of potential cerebrovascular disease leading to prolonged ATT, which is associated with arterial stiffness, and vascular resistance. The clusters demonstrating the association between FRS and ATT are the brain regions that may have heightened vulnerability to vascular risk factors, according to previous studies.
Unobtrusive In‐Home Digital Monitoring of Frontotemporal Dementia: Feasibility and Adherence
AbstractBackgroundFrontotemporal Dementia (FTD) patients are geographically dispersed and often have behavioral and/or motor deficits that limit access to in‐person clinical trials and research. To address these participation barriers, we investigated the feasibility of installing in‐home sensor technologies to remotely assess everyday functioning in FTD.MethodWe aimed to recruit 20 FTD participants with a variety of clinical syndromes, including behavioral variant FTD (bvFTD), primary progressive aphasia (PPA), corticobasal syndrome (CBS), and progressive supranuclear palsy (PSP). We also aimed to enroll a study partner for each participant that lived in the same home to serve as a control. The Collaborative Aging Research Using Technology (CART) platform was installed in participants’ homes to collect continuous data for up to two years, using motion sensors, physiological monitoring devices (e.g., scales, sleep sensors), electronic pillboxes, and wearable actigraphs. Computer use and driving patterns were also tracked. Technology was installed either via a study coordinator visit or remotely through video and written instructions, based on schedule and logistics. Participants were asked to complete a weekly Qualtrics survey about time away from home, visitors, and health changes.ResultOf the 159 UCSF participants available for recruitment through ALLFTD and other FTD projects, 28% (N = 45) met inclusion criteria. Of those 45, ten were ineligible due to advanced disease severity or relocation to an assisted living facility. A third (36%, N = 17) declined due to overwhelmed study partners (N = 3), declined any subproject participation (N = 7), sensitivity to participant’s diagnosis (N = 2), or lack of response (N = 5). Two additional participants were recruited through UCLA’s ALLFTD cohort. The platform has been installed in 19 homes (bvFTD: n = 10; PPA: n = 5; CBS/PSP: n = 4). Two homes discontinued due to death or relocation to assisted living facility. Mean data collection is 360 days (max = 527). Of 103 reported technological issues, 92% were resolved without additional home visits, either by phone or email. Participants completed 65% of weekly health questionnaires.ConclusionsAlthough there were some recruitment and retention challenges, these findings generally provide preliminary support for the feasibility and sustainability of conducting comprehensive in‐home passive data research in FTD.
Changes in cortical thickness along the Alzheimer disease continuum in people with Down syndrome
AbstractBackgroundTriplication of the APP allele in Down syndrome (DS) leads to excess amyloid production and Alzheimer’s disease (AD) related cognitive decline. Key biomarkers (amyloid, tau, neurodegeneration) can identify pathological processes that occur before clinical symptoms and more precisely stage adults with DS along the AD continuum. Previous research has used changes in cortical thickness (CT) as an indicator of neurodegeneration in autosomal‐dominant AD1, but not DS‐related AD. It is unclear when and where cortical thinning occurs along the AD continuum in DS and if these differences are associated with increasing AD pathology.MethodCross‐sectional analysis compared DS participants from the Alzheimer Biomarkers Consortium‐Down Syndrome (106 amyloid‐/cognitively unimpaired, nonclinical, 20% reserved for receiver‐operating characteristic, ROC, analyses; 45 amyloid+/cognitively unimpaired, preclinical; 27 amyloid+/cognitively impaired, clinical) who had amyloid positron emission tomography and magnetic resonance imaging (MRI). Amyloid positivity was defined using standardized uptake value ratio (partial volume corrected) cortical mean values (1.42/1.19 for PIB/AV452) and impairment was based on consensus diagnosis. CT differences between nonclinical and clinical groups were calculated in FreeSurfer v5.3 for multiple thresholds (p<.05/.01/.005/.001/.0005/.0001) using a cluster‐wise false discovery rate of .001, corrected across hemispheres, after controlling for age and sex. CT values were extracted for each threshold and used in ROC analyses (nonclinical reserved participants compared against preclinical and preclinical compared against clinical) to identify key cortical regions that best discriminated amyloid positivity and between unimpaired and impaired.ResultSignificant decreases in parietal and inferior temporal CT were observed between nonclinical and clinical (Figures 1 and 2). ROC analyses identified regions in the left (threshold .05) and right (threshold .005) hemispheres that best differentiated nonclinical from preclinical. Observed area under the curves (AUCs) of .772 and .830 exceeded those previously reported for autosomal‐dominant AD (.547/.540)1. Analyses differentiating preclinical from clinical observed AUCs of .786 for left (threshold .0005) and .770 for right (threshold .0005) hemispheres.ConclusionCT differences primarily within posterior cortical regions differentiated adults with DS by amyloid status and cognitive impairment. CT differences were more diffuse for amyloid positivity but more focal for impairment.1. Dincer (2020). NeuroImage: Clinical, 28, 102491.2. Su (2019). Alzheimer’s Dement, 11, 180‐190.
Flortaucipir tau PET in former American football players
AbstractBackgroundChronic traumatic encephalopathy (CTE) is a neurodegenerative disease thought to be caused, in part, by repetitive head impact (RHI) exposure and characterized by phosphorylated tau deposition with varying patterns in the cerebral cortex. Flortaucipir tau PET data from the DIAGNOSE CTE Research Project was examined to investigate tau PET patterns and its relation to RHI exposure in former professional and college American football players.MethodIn this study, 218 DIAGNOSE CTE participants, ages 45‐74, with flortaucipir PET were included; 104 were former National Football League players (PRO), 58 were former college football players (COL), and 56 had no history of contact sports, other exposure to RHI, or traumatic brain injury (UE). Regional flortaucipir uptake was quantified using established procedures for three previously defined statistical regions (superior frontal, left parietal, and medial temporal) and compared across the three groups. Association analyses were also performed to examine the relationship between flortaucipir uptake and exposure to RHI using the cumulative head impact index for linear acceleration g‐force (CHII‐G). The potential clinical utility of flortaucipir PET for CTE was also investigated by examining its ability to differentiate participants meeting criteria for traumatic encephalopathy syndrome (TES) and probable or possible level of certainty for CTE pathology from those who do not meet those criteria.ResultSignificantly higher flortaucipir uptake was observed in former football players than in participants without RHI exposure, controlling for age and race for the superior frontal region (Fig. 1A), left parietal, and medial temporal regions. Association between regional flortaucipir uptake and CHII‐G was only observed in the superior frontal region in former players (PRO+COL) 60 years and older (Fig. 1B) and no association was observed in the full PRO+COL group. Flortaucipir PET was not able to differentiate TES groups in former football players (Fig. 1C).ConclusionElevated flortaucipir uptake was observed in former American football players to a group of unexposed men. The association between flortaucipir uptake and exposure to RHI may require a long latent period to manifest. Further investigation of more sensitive tau PET tracers and quantification techniques is warranted to better characterize CTE related tau pathology.
Cohort heterogeneity and AD biomarkers in older adults without significant cognitive impairment: A comparison of U.S. POINTER and ADNI
AbstractBackgroundA key focus of the U.S. Study to Protect Brain Health Through Lifestyle Intervention to Reduce Risk (U.S. POINTER) is recruitment of unimpaired older individuals at risk for cognitive decline with heterogeneous characteristics (under‐represented racial and ethnicity groups (URG), family history of cognitive decline, hypertension, geographic diversity), and a subset of individuals undergo PET and MRI.MethodTo examine effects of heterogeneity on relationships between risk factors, AD biomarkers (Aβ and entorhinal tau PET, hippocampal volume (HV)) and cognition (global, executive function, and memory) were examined from U.S. POINTER neuroimaging participants (N = 602, about 65% of the anticipated total sample) and a subset of Alzheimer’s Disease Neuroimaging Initiative participants (N = 503). Neuroimaging data were processed and analyzed using harmonized methods across studies. Some cohort‐specific measurements were standardized using a within‐cohort reference group (Aβ‐negative individuals <70yrs and with global CDR = 0).ResultU.S. POINTER imaging participants are younger than those in ADNI (68.6+/‐5.0yrs vs 70.7+/‐4.4yrs), more likely to be female (65% vs 58%) and from URG (27% vs 15%), and had lower education (14.7+/‐3.3yrs vs 16.7+/‐2.3yrs) and higher Framingham Cardiovascular Risk scores (9.9+/‐8.3 vs 7.7+/‐5.3). The groups differed on several AD biomarkers and cognitive measures (Table). In regression models, AD biomarkers were stronger predictors of cognitive performance in ADNI compared to POINTER. Specifically, Aβ+ individuals with elevated tau had disproportionately poorer cognitive performance in ADNI, and this relationship remained consistent across cognitive measures (Figure). Conversely, demographic/risk variables were stronger predictors in POINTER. Demographic/risk predictors (age, sex, education, race, and FRS) together explained twice as much variance in global cognition in POINTER compared to ADNI (Adjusted R2 = 29% vs 14%). The addition of AD biomarkers to the model further increased the amount of variance in global cognition explained from 14% to 34% in ADNI, but addition of AD biomarkers did not increase variance explained in the POINTER model.ConclusionHeterogeneity among study participants appears to influence baseline biomarker‐cognition relationships such that AD biomarkers explain more variability in cognitive performance in ADNI than POINTER. These findings further suggest that cohort characteristics may affect the potential influence of amyloid‐ and tau‐modifying treatments on cognition.
Harmonization of amyloid and tau PET tracers in a multi‐center cohort from NIA Alzheimer’s disease research centers in the United States
AbstractBackgroundThe Standardized Centralized Alzheimer’s & related dementias Neuroimaging (SCAN) project harmonizes processing of prospectively acquired PET and MRI scans from Alzheimer’s disease research centers (ADRCs) across the USA. SCAN data will be integrated with other ADRC data streams within the National Alzheimer’s Coordinating Center (NACC), including the Uniform Data Set which includes rich longitudinal demographic, cognitive, and neuropathological data, resulting in a large, potentially representative cohort. The project has sought to accommodate site‐to‐site variation in image acquisition while permitting harmonization of imaging data, necessitating new analytic approaches for different PET tracers and PET scans without MRIs.MethodPET scans underwent preprocessing for quality control and smoothing to a uniform resolution. MRI‐free processing pipelines and positivity thresholds were developed for each beta‐amyloid (Aβ)‐PET tracer: [11C]PiB, [18F]florbetaben (FBB), [18F]florbetapir (FBP) and [18F]flutafuranol (NAV). A generic Aβ‐PET template was used for spatial normalization and cortical summary SUVRs were calculated in template space. Aβ‐positivity thresholds were determined by examining both MRI‐dependent thresholds transformed into MRI‐free units and a young control mean+2SD. Centiloid (CL) conversions were completed for each Aβ‐PET tracer. Tau‐PET tracers ([18F]flortaucipir (FTP) and [18F]MK‐6240) required tracer‐specific warping templates for best performance relative to gold‐standard MRI‐dependent processing. Demographic and clinical data were obtained from NACC.Result15 ADRCs have contributed data to SCAN as of January 2023. 477 participants had preprocessed, SCAN‐compliant (https://scan.naccdata.org/) Aβ‐PET scans and NACC data; 219 of these participants (46%) also had tau‐PET scans (Table 1). This cohort included 18% of individuals from underrepresented groups. Clinical diagnostic groups ranged from cognitively unimpaired to MCI and dementia (74% of whom had AD dementia). Using pooled Aβ‐PET data across 4 tracers in CLs, there were expected differences in Aβ burden by diagnosis group (Fig. 1) and positive associations with tau (Fig. 2).ConclusionSCAN has successfully harmonized PET data from more than a dozen US ADRCs yielding a large, clinically heterogeneous cohort of deeply phenotyped research participants. SCAN data will be compatible with ADNI data allowing for further pooling of data and greater statistical power to drive critical questions in AD and related dementias research.
Identification of genes and blood co‐expression networks associated with cerebrovascular neuroimaging phenotypes
AbstractBackgroundA hallmark of Alzheimer’s disease is amyloid‐β (Aβ) deposition in the brain parenchyma as senile plaques which Aß frequently accompanies in the cerebral vasculature as cerebral amyloid angiopathy (CAA). Previous studies demonstrated that the cerebrovascular system is essential for Aβ clearance and interactions exist between cerebrovascular disease and AD. Neuroimaging data, including measures on small vessel disease, amyloid deposits, and microhemorrhages (MCH), can reflect both cerebrovascular disease and AD pathology, thereby serving as intermediate phenotypes of both conditions. We aim to identify genes and co‐expression networks in blood associated with these neuroimaging endophenotypes and the biological pathways enriched in these perturbed transcriptional networks.MethodWe conducted differential gene expression and Weighted Gene Co‐Expression Network analyses (WGCNA) on blood RNA sequencing data of 407 participants from the longitudinally followed Mayo Clinic Study of Aging cohort. The neuroimaging phenotypes include MCH, infarction, and amyloid deposits measured by Pittsburgh compound B positron emission tomography (PiB PET).ResultWGCNA analysis showed a nominally significant association (p<0.05) between several module eigengenes (MEs) and MCH and PiB PET. One module, M42, remained significantly associated with PiB PET after Bonferroni correction for all 43 modules tested. Pathway enrichment analysis of M42 identified two significant pathways (FDR‐adjusted p<0.05) including phosphodiesterase I activity and IgE binding. These pathways contain genes such as MS4A2 encoding the high‐affinity IgE receptor involved in the immune response. Similarly, ENPP2 in the phosphodiesterase I activity pathway represents an additional candidate gene previously implicated in Alzheimer’s disease.ConclusionWe generated co‐expression networks and identified one network module with a statistically significant association between brain amyloid deposits and blood transcriptome data. Further investigations of this network and its enriched pathways are necessary to understand their roles in amyloidosis, CAA, and neurodegeneration. Future studies will include seeking replication of these results in other datasets. Such studies can help elucidate the biological mechanisms of pathogenesis of cerebrovascular damage and amyloid deposits. In addition, in a clinical setting, blood levels of transcripts discovered in this study may serve as predictive or diagnostic biomarkers with mechanistic implications.
White matter hyperintensities: A long‐term consequence of repetitive head impacts?
AbstractBackgroundRepetitive head impacts (RHI) from American football can lead to tau and non‐tau pathologies that might present as white matter hyperintensities (WMH) on FLAIR MRI. In 2022, we published a study in Alzheimer’s & Dementia that examined WMH and their association with risk factors and clinical function in former elite football players. That study will be presented during this symposium along with new unpublished data that used imaging and fluid biomarkers to examine etiological correlates (tau, amyloid, neurodegeneration, axonal injury, neuroinflammation) of WMH in the same cohort.MethodsThe multi‐site DIAGNOSE CTE Research Project recruited 180 football players and 60 asymptomatic males without RHI (“controls”), all 45‐74 years (Table 1). Participants completed MRI, lumbar puncture, and neuropsychological testing. Lesion Segmentation Toolbox estimated FLAIR WMH, FreeSurfer derived total cortical thickness, and a diffusion pipeline derived average global FA. CSF was analyzed for p‐tau181, Aβ1‐42, NfL, and sTREM2. Tobit regressions compared football players (n = 149) and controls (n = 53) on total and regional log‐WMH and estimated the effects of years of football and age of first exposure (AFE) to football on log‐WMH. Linear regressions evaluated log‐WMH and clinical associations in football players. Structural equation modeling (SEM) examined effects between log‐WMH and cortical thickness, FA, and the CSF biomarkers. Analyses accounted for age, race, revised Framingham Stroke Risk Profile (rFSRP), body mass index, APOE ε4, and evaluation site.ResultsAlosco et al. (2022) found older (i.e., 60+) but not younger football players had greater total, frontal, temporal and parietal log‐WMH compared to controls (FDR‐adjusted p‐values<0.05; Table 2). Among older football players, younger AFE was associated with greater log‐WMH (beta = ‐0.13, 95% CI = ‐0.23,‐0.02). Greater log‐WMH corresponded to worse Trails A‐B (beta = ‐4.32, 95% CI = ‐8.59,‐0.05) and List Learning Long Delay scores (beta = ‐0.56, 95% CI = ‐0.93,‐0.19). Subsequent unpublished SEM analyses in football players (Figure 1) showed the following direct effects on log‐WMH: higher rFSRP, higher CSF p‐tau181, lower FA, and reduced cortical thickness. There were no indirect effects. Compared to controls, SEM associations were stronger in football players exception for cortical thickness.ConclusionsFLAIR WMH might have unique imaging characteristics, risk factors, and pathological underpinnings in people exposed to RHI.
Baseline and Six‐Month Actigraphy and Sleep‐monitoring Devices Help Differentiate Between Alzheimer’s Disease and Dementia with Lewy Bodies
AbstractBackgroundDementia with Lewy bodies (DLB) and Alzheimer’s disease (AD) are challenging to differentiate. Mobile health devices that measure motor and sleep domains between clinic visits may improve differential diagnosis.MethodParticipants with probable AD and probable DLB wore ActiGraph triaxial wGT3X‐BT actigraphy devices on their nondominant wrists for 2 weeks and multichannel Sleep ProfilerX8TM EEG devices on their foreheads for 2 consecutive nights.3 Actigraphy data were processed and scored using GGIR and activity counts in R. The Sleep Profiler characterizes abnormal slow‐wave sleep and non‐REM hypertonia (NRH).ResultAt baseline, participants with DLB (n = 9) had higher UPDRS scores and a higher likelihood of psychosis than participants with AD (n = 8; see Table 1). Using univariate Wilcoxon Rank‐Sum tests, we found that at baseline participants with DLB had lower daily activity counts, higher % time in NRH, and lower % time in REM sleep than participants with AD (unadjusted p‐values < 0.05; See Figure 1).In a recent 6‐month followup, activity levels were lower on average than baseline activity levels for both DLB (n = 5) and AD (n = 5). at 6‐months, the differences in % sleep in NRH increased between DLB participants (n = 5) and AD participants (n = 3) as average % NRH increase 6% in DLB and decreased 1% in AD (see Table 2). These 6‐month differences are not statistically significant, but data collection are ongoing.An exploration of methods for processing and scoring actigraphy data revealed that processing can result in wide range of findings. As a result, we have focused on actigraphy outcomes which are more suitable to older participants with dementia.ConclusionParticipants with DLB demonstrated lower baseline physical activity and higher rates of baseline sleep disturbances than participants with AD; sleep differences between DLB and AD appeared to increase over 6 months.Actigraphy and sleep‐monitoring devices are well tolerated in participants with dementia and they can reliably collect valid, useful activity and sleep data in between clinic visits.
Targeted plasma eQTL and pQTL to identify potential novel biomarkers for Alzheimer’s Disease in African Americans
AbstractBackgroundGiven the overlap of symptoms between Alzheimer’s disease (AD) and other dementias, it is essential to develop widely accessible biomarkers that will ensure an accurate diagnosis of AD to enable its proper treatment. African Americans (AA) are twice more likely to develop AD than non‐Hispanic whites (NHW) and yet they remain underrepresented in AD research. The few studies conducted in AA suggest that there are different factors that influence AD in this population; therefore, it is possible that different biomarkers and/or therapies may be needed to effectively treat AD in AA. In this study we are testing the hypothesis that genetic variants that show association with AD‐risk in AA and with plasma levels of transcript and proteins may serve as AD biomarkers which are more accessible and less costly than cerebrospinal fluid or neuroimaging biomarkers.MethodsIn 189 AD cases and 183 cognitively unimpaired (CU) AA participants of the Florida Consortium of African American Alzheimer’s Disease Studies (FCA3DS), we performed targeted DNA sequencing of 10 AD‐associated loci. Genetic variants were tested for association with plasma transcripts and plasma total tau protein levels previously measured in this cohort using a nanoString® custom panel (Reddy J, et al. 2022) and Simoa assays (Deniz K, et al. 2021), respectively. In receiver operating characteristic (ROC) analyses, the most significant expression/protein quantitative trait loci (eQTL/pQTL) were added sequentially to a base model that included only age and sex to identify eQTL/pQTL that improved the accuracy to correctly classify AD cases vs. CU.ResultsOf the 5112 variants identified in the targeted sequencing, 71 were nominally associated (p<0.05) with plasma transcript/protein levels and with AD‐risk. ROC analysis of age, sex, eQTL/pQTL in APOE, ABCA7, AKAP9, EPHA1, SORL1, and TLR4, and plasma levels of APP, ABCA7, AKAP9, CD14, CLU, and APOE‐ɛ4 dosage achieved 77.6% area under the curve (AUC) to discriminate AD vs. CU, a 19.9% improvement over a model that only included age and sex.ConclusionsPlasma transcript levels and eQTL/pQTL are promising diagnostic biomarkers for AD that may improve accessibility and reduce costs for a more accurate diagnosis of AD.