Windham, B. Gwen

Abstract Background Prospective investigations on blood biomarkers for Alzheimer's disease and related‐dementias (ADRDs) and cognitive impairment provide important insights into their clinical utility in early disease diagnosis and risk stratification, but have been primarily limited to AD pathologies (e.g. amyloid‐beta and hyperphosphorylated‐tau aggregation). Studies of biomarkers reflecting other pathogenic processes that contribute to ADRDs, especially in community‐based cohorts, are lacking. We examined associations of blood biomarkers reflecting glial and neuronal dysfunction and neuroinflammation, including soluble cluster of differentiation‐14 (sCD14), chitinase‐3‐like protein‐1 (YKL‐40), glial‐fibrillary acidic protein (GFAP), neurofilament‐light chain (NfL), total‐Tau, and Ubiquitin C‐terminal hydrolase L1 (UCH‐L1) with cognitive decline and incident dementia in ARIC study participants. Methods Blood biomarkers were measured in 1,864 ARIC Visit 3 (1993‐1995) participants using the Meso Scale Discovery assay (YKL‐40), the Quanterix Simoa Neurology 4‐Plex assay (total‐Tau, NfL, GFAP, UCH‐L1), and a commercial ELISA assay (sCD14, R&D Systems); global cognition z‐scores were derived from three cognitive tests (measuring memory, language, and executive function) administered up to six times over 25 years, and incident dementia cases were ascertained via in‐person cognitive assessments (expert‐adjudicated), informant‐based interviews, and hospitalization and death records through 2018. We used shared parameter models to simultaneously fit linear mixed‐effect models and proportional hazard models to estimate 25‐year cognitive decline and dementia rates separately for each biomarker (log‐base2‐transformed), adjusting for covariates (Table). Results Among 1,864 participants (mean±SD age:63±4.5 years, 60% female, 49% self‐reported Black), 27% developed dementia. Higher levels of each biomarker were associated with faster cognitive decline. All but UCH‐L1 were associated with elevated dementia risk (Table). For example, one standard deviation (SD) higher log‐base2 YKL‐40 was associated with ‐0.10 SD steeper 25‐year cognitive decline (95%CI:‐0.15,‐0.04) and 59% higher dementia hazard (hazard ratio (HR):1.59 (95%CI:1.20,2.10); likewise, sCD14 was associated with an additional cognitive decline of ‐0.06 (95%CI:‐0.11,‐0.01) and 38% higher dementia hazard (HR:1.38 (95%CI:1.08,1.76)). Conclusions Blood biomarkers reflecting neuropathological processes, including microglial activation, astrocytic reactivity, and neuronal degeneration, were associated with faster cognitive decline and increased dementia risk, suggesting that they may aid in risk stratification and intervention strategies for ADRDs. Future analyses will examine whether these associations differ by race and APOE‐ε4 s tatus.

Abstract Background Even among cognitively unimpaired (CU) older adults, specific patterns of neuropsychological performance impart differential risk for progression to Alzheimer's disease and related dementias (ADRD). Early identification in the prodromal phase may inform early intervention efforts. We used a data‐driven approach to classify neuropsychological phenotypes within CU older adults from the Atherosclerosis Risk in Communities (ARIC) study and tested for differences in ADRD and cerebrovascular outcomes. Method Participants included 4737 CU (via a previously published algorithm plus consensus diagnosis protocol) older adults (mean age=75.1 years, 60.8% female, 23.1% Black) with neuropsychological data from Visit 5 (2011‐2013) of the ARIC study. Latent profile analysis was leveraged to identify unique cognitive phenotypes using 10 demographically adjusted (age, sex/gender, education, race‐center, WRAT Reading) neuropsychological z‐scores. Cognitive phenotypes were then compared on demographics, depressive symptoms, and vascular risk factors (VRFs). In a subset with available neuroimaging and biofluid data ( n range=1102‐1678), white matter hyperintensity (WMH) and bilateral hippocampal (HV) volumes normalized for intracranial volume and plasma biomarkers (i.e., Aβ42/40, ptau181, NfL, and GFAP) were examined by phenotype. Phenotype comparisons adjusted for demographic variables, VRFs, and an indicator of kidney function (i.e., eGFR‐creatinine) for plasma biomarkers. We applied Tukey adjustment for pairwise comparisons. Result A 6‐class solution of cognitive phenotypes emerged as the best fitting model: Average Overall (28%), Low Overall/Very Low Memory (7%), Average Overall/Super Memory (11%), Subtle Low Overall/Low Executive (22%), Subtle Low Executive/High Memory (13%), and Average Overall/Low Memory (19%; Figure 1). Phenotypes differed on education, race, and word reading, despite prior adjustments for these variables when creating the neuropsychological z‐scores (Table 1). Additionally, phenotypes differed on depressive symptoms, VRFs, WMH, HV, Aβ42/40, and ptau181 (all p s < .05). The Subtle Low Overall/Low Executive group had the highest WMH volumes, the Subtle Low Executive/High Memory group had the lowest Aβ42/40, and the Low Overall/Very Low Memory group had the lowest HV and highest levels of plasma ptau181 (Table 2). Conclusion Cognitive heterogeneity is observed even among CU older adults, with lower executive performance associated with cerebrovascular changes and lower memory associated with certain AD‐related changes, possibly indicating risk for cognitive progression through distinct pathways.

Abstract Background Diffusion MRI (dMRI) measures are increasingly being used as biomarkers for aging and dementia in multi‐center studies. Measurements in longitudinal studies are often acquired on different systems, which can result in systematic differences of machine origin (e.g. vendor, hardware, algorithms). This may warrant harmonization prior to pooling multi‐site data. However, harmonization algorithms may introduce a risk of removing valuable information content and reducing sensitivity to physiological factors (e.g. sex, age, disease). Here, we tested the hypothesis that protocol standardization may minimize machine‐origin differences, obviating a need for post‐acquisition harmonization in a two‐site (Mississippi‐MS and Minnesota‐MN) cohort of the longitudinal Mayo Clinic Study of Aging (MCSA). Methods Two data sets were analyzed: first, a crossover dataset with 80 participants (scanned within 2 days) bridging two different protocols on GE and Siemens scanners; second, participants scanned using standardized protocols from the ongoing MCSA ( N MN‐MCSA =204; N MS‐MCSA =52). Parameters are listed in Figure 1. For crossover data, analyses were repeated with GE data resampled to 2‐mm isotropic resolution, and discarding Siemens‐only b‐shells to artificially reduce protocol disparities. Analysis encompassed visualization of overlain subjectwise histogram‐derived densities obtained from whole‐brain masks for dMRI measures; and median dMRI measures from within the genu of the corpus callosum (GCC) with nonparametric statistical testing. When available, both DTI and NODDI measures were compared. Results Pronounced visual distributional differences seen between MD and FA in crossover data were reduced with resampling of GE data and including only b‐shells shared across acquisitions. For the resampled/reshelled data, significant differences remained for all GCC‐FA comparisons, but not all GCC‐MD (Figure 2). For the ongoing MCSA data from two sites, MD and FA whole‐brain densities were more closely aligned, with GCC‐FA and GCC‐MD showing no significant differences (Figure 3). NODDI measures showed more discordant densities than DTI measures, including shifts in most probable values and differences in distributional arrangement. Conclusions Our work provides further evidence supporting protocol and processing standardization for minimizing machine‐origin differences prior to considering harmonization. Differences in dMRI metrics seen when comparing across unstandardized protocols were reduced when protocols were standardized. NODDI metrics had more apparent sensitivity to site than DTI measures.

Abstract Background Diffusion MRI (dMRI) measures are increasingly being used as biomarkers for aging and dementia in multi‐center studies. Measurements in longitudinal studies are often acquired on different systems, which can result in systematic differences of machine origin (e.g. vendor, hardware, algorithms). This may warrant harmonization prior to pooling multi‐site data. However, harmonization algorithms may introduce a risk of removing valuable information content and reducing sensitivity to physiological factors (e.g. sex, age, disease). Here, we tested the hypothesis that protocol standardization may minimize machine‐origin differences, obviating a need for post‐acquisition harmonization in a two‐site (Mississippi‐MS and Minnesota‐MN) cohort of the longitudinal Mayo Clinic Study of Aging (MCSA). Methods Two data sets were analyzed: first, a crossover dataset with 80 participants (scanned within 2 days) bridging two different protocols on GE and Siemens scanners; second, participants scanned using standardized protocols from the ongoing MCSA ( N MN‐MCSA =204; N MS‐MCSA =52). Parameters are listed in Figure 1. For crossover data, analyses were repeated with GE data resampled to 2‐mm isotropic resolution, and discarding Siemens‐only b‐shells to artificially reduce protocol disparities. Analysis encompassed visualization of overlain subjectwise histogram‐derived densities obtained from whole‐brain masks for dMRI measures; and median dMRI measures from within the genu of the corpus callosum (GCC) with nonparametric statistical testing. When available, both DTI and NODDI measures were compared. Results Pronounced visual distributional differences seen between MD and FA in crossover data were reduced with resampling of GE data and including only b‐shells shared across acquisitions. For the resampled/reshelled data, significant differences remained for all GCC‐FA comparisons, but not all GCC‐MD (Figure 2). For the ongoing MCSA data from two sites, MD and FA whole‐brain densities were more closely aligned, with GCC‐FA and GCC‐MD showing no significant differences (Figure 3). NODDI measures showed more discordant densities than DTI measures, including shifts in most probable values and differences in distributional arrangement. Conclusions Our work provides further evidence supporting protocol and processing standardization for minimizing machine‐origin differences prior to considering harmonization. Differences in dMRI metrics seen when comparing across unstandardized protocols were reduced when protocols were standardized. NODDI metrics had more apparent sensitivity to site than DTI measures.

Background: Chronic kidney disease (CKD) has been linked to elevated inflammatory biomarkers, yet few studies have evaluated these associations using iothalamate-based measures of kidney function, highlighting the need for further research. Research question: Is CKD associated with elevated inflammatory markers when defined by measured glomerular filtration rate (mGFR) and urinary albumin-to-creatinine ratio (UACR)? Methods: Participants from the Genetic Epidemiology Network of Arteriopathy (GENOA) cohort were included if they had data available for at least one inflammatory marker at visit 2. 17 inflammatory markers were assessed, including C-reactive protein (CRP), interleukin-6 (IL-6), IL-18, tumor necrosis factor receptors 1 and 2 (TNFR1, TNFR2), serum amyloid A, intercellular and vascular adhesion molecules (ICAM, VCAM), E-selectin, P-selectin, monocyte chemoattractant protein (MCP), myeloperoxidase (MPO), receptor for advanced glycation end products (RAGE), matrix metalloproteinases 2 and 9 (MMP-2, MMP-9), and tissue inhibitors of metalloproteinases 1 and 2 (TIMP-1, TIMP-2). Kindey function was assessed using the urinary clearance of iothalamate. CKD was defined as mGFR < 60 mL/min/1.73 m 2 or UACR ≥ 30 mg/g. To further assess whether associations persisted without albuminuria, we also defined CKD solely by mGFR < 60 mL/min/1.73 m 2 . Generalized estimating equations (GEE) were used to assess associations between CKD status and inflammatory markers, accounting for familial clustering. Exponentiated beta coefficients from GEE models were used to present geometric mean (GM) ratios of each biomarker by CKD status, adjusted for demographics, lifestyle factors, cardiovascular risk factors, and medication use. Results: Among 1,010 participants (mean age 61 ± 10 years; 67% women; 48% Black adults), 207 were classified as having CKD. After adjusting for potential confounders, individuals with CKD had significantly higher levels of hsCRP, VCAM, TNFR1, and TNFR2 compared to those without CKD (GM ratios > 1; p < 0.05), while the remaining inflammatory markers showed no significant differences ( Figure 1 ). Importantly, these elevations remained significant when CKD was defined by mGFR alone ( Figure 2 ). Conclusion: Our findings suggest that CKD, defined by mGFR, is associated with specific inflammatory biomarkers, underscoring the need for mechanistic studies to investigate these associations.

AbstractINTRODUCTIONWe investigated the usefulness of plasma neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) for capturing vascular cognitive impairment (VCI) in the context of amyloidosis.METHODSUsing two independent cohorts (n = 1810), we assessed the relationship of plasma NfL and GFAP with (1) vascular brain indices; (2) diagnostic states using the following definitions: vascular versus not (white matter hyperintensity/total intracranial volume ≥ 1.3%), and cognitively impaired (CI) versus cognitively unimpaired (CU) using Clinical Dementia Rating ([CDR] scale ≥ 0.5); and (3) their upstream predictors using structural equation models (SEMs).RESULTSPlasma NfL and GFAP were associated with vascular brain damage and differed across states (VCI > vascular CU > non‐vascular CI > non‐vascular CU). In a population‐based sample, these biomarkers distinguished vascular CU and VCI from non‐vascular CU groups with greater separation in amyloid negative participants. Pathway analyses showed NfL was primarily influenced by systemic/brain vascular health, whereas amyloid contributed to GFAP variability.DISCUSSIONPlasma biomarkers, particularly NfL, capture vascular brain changes and show promise for VCI identification.Highlights Plasma NfL and glial fibrillary acidic protein (GFAP) were associated with vascular brain indices. Plasma biomarkers differed across diagnostic states (VCI > vascular CU > non‐vascular CI > non‐vascular CU). Plasma markers discriminated vascular from non‐vascular states with greater separation in Aβ− participants. NfL was linked to vascular health, while amyloid influenced GFAP variability in the population‐based sample. Future longitudinal frameworks should consider systemic inflammation markers along with these plasma markers to better understand VCID‐related brain changes and cognitive decline.

ImportanceMidlife vascular risk factors are associated with an elevated risk of dementia. However, the total contribution of vascular risk factors in midlife and late life with incident dementia is uncertain.ObjectiveTo quantify the proportion of incident dementia attributable to modifiable vascular risk factors measured in midlife and late life and to examine differences by apolipoprotein ε4 genotype, self-reported race, and sex.Design, Setting, and ParticipantsThis was a prospective cohort analysis of the Atherosclerosis Risk in Communities (ARIC) study using 33 years of follow-up (1987-2020). The setting included ARIC field centers (Jackson, Mississippi; Forsyth County, North Carolina; Minneapolis suburbs, Minnesota; Washington County, Maryland). Study baseline in Black and White participants with complete exposure and covariate data was set by age at risk factor measurement (45-54 years, 55-64 years, and 65-74 years). Data were analyzed from August 2023 to December 2024.ExposuresHypertension (systolic blood pressure [BP] ≥130 mm Hg, diastolic BP ≥80 mm Hg, or use of medication for BP), diabetes (fasting glucose ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, self-reported physician’s diagnosis, or use of any diabetes medication), and current smoking (self-reported).Main Outcomes and MeasuresIncident dementia. Population attributable fractions were estimated by age 80 years, and separately after 80 years, from having at least 1 vascular risk factor by age at risk factor measurement.ResultsA total of 7731 participants were included in analysis of risk factors measured at age 45 to 54 years (4494 female [58%]; 2207 Black [29%]; 5524 White [71%]), 12 274 contributed to analysis of risk factors measured at age 55 to 64 years (6698 female [55%]; 2886 Black [24%]; 9388 White [76%]), and 6787 contributed to analysis of risk factors measured at age 65 to 74 years (3764 female [56%], 1375 Black [20%]; 5412 White [80%]). There were 801, 995, and 422 dementia cases by 80 years, respectively. The fraction of dementia by 80 years attributable to at least 1 vascular factor at age 45 to 54 years was 21.8% (95% CI, 14.3%-29.3%), at 55 to 64 years was 26.4% (95% CI, 19.1%-33.6%), and at 65 to 74 years was 44.0% (95% CI, 30.9%-57.2%). Attributable fractions for these factors were higher in apolipoprotein ε4 noncarriers at age 55 years and older (range, 33.3%-61.4%), Black individuals at age 45 years and older (range, 25.5%-52.9%), and female individuals at age 55 years and older (range, 29.2%-51.3%). Only 2% to 8% of dementia cases after 80 years were attributable to these factors.Conclusions and RelevanceResults of this cohort study suggest that between 22% and 44% of incident dementia cases by 80 years in the ARIC study were attributed to midlife and late-life vascular risk factors. Assuming causal relationships, maintaining optimal vascular health across the life course could mitigate a sizeable proportion of dementia risk by 80 years.

Introduction: Physical activity (PA) has been identified as a protective factor for dementia, but the mechanisms linking PA to brain health are unclear. To elucidate potential pathways linking PA to brain health, we tested the hypothesis that higher levels of late-life PA are associated with a lower burden of Alzheimer’s disease (AD) pathology, cerebrovascular disease, and neurodegeneration measured through brain imaging and plasma biomarkers. Methods: ARIC-NCS participants without dementia in 2011-13 and with measures of PA, 3T brain magnetic resonance imaging (MRI), and stored plasma samples were examined (n=1,771; mean age: 76.5 years; 40.5% male; 27.6% Black). Participants were classified as meeting or not meeting 2018 aerobic PA guidelines based on a standardized interviewer-administered questionnaire. MRI biomarkers included white matter hyperintensity (WMH) volume (cm 3 ), total brain volume (TBV) (cm 3 ), and cortical thickness (CT) (mm). Plasma biomarkers of the ratio of amyloid beta 42 to 40 (Aβ42/Aβ40), phosphorylated Tau-181 (p-Tau181), neurofilament light chain (NfL), and glial fibrillary acidic protein (GFAP) were measured using Quanterix Simoa immunoassays. All biomarkers were standardized to measures in 2011-13 to facilitate comparisons. Covariate-adjusted linear regression models estimated cross-sectional associations of PA with MRI or plasma biomarkers. Covariate-adjusted linear mixed effects models quantified associations of PA with the 5-year rate of change in a subset of participants with MRI (n=669) or plasma (n=651) biomarkers at both baseline and follow-up from 2011-19 (median follow-up: 5.4 years). Results: Meeting versus not meeting aerobic PA guidelines was cross-sectionally associated with lower volumes of WMH (β = -0.16, 95% CI: -0.25, -0.07), higher TBV (β = 0.09; 95% CI: 0.05, 0.13), and lower plasma NfL (β = -0.102, 95% CI: -0.184, -0.021) ( Figure ). Late-life PA was not associated with 5-year rates of change in MRI or plasma biomarkers. Conclusions: Late-life PA was cross-sectionally associated with less neurodegeneration and cerebrovascular disease. Associations with 5-year changes in imaging and plasma biomarkers of AD pathology, cerebrovascular disease, or neurodegeneration were not statistically supported. Longitudinal biomarker studies with wearable devices that quantify other daily intensity-based movement metrics (e.g., light PA) and larger sample sizes are needed to further examine mechanisms linking PA to brain health.

Introduction: Physical activity (PA) has been identified as a modifiable factor inversely associated with cognitive decline and dementia risk. However, most longitudinal studies to date have short follow-up and report inconsistent results on the benefit of PA to support brain health among older adults. In this study, we hypothesized that higher levels of PA measured in late life would be associated with slower rates of global and domain-specific cognitive decline and a lower risk of dementia over ~10 years in late life. Methods: Dementia-free participants (n=5,982; mean age: 76 years; 42% male, 23% Black) who completed the interviewer-administered Modified Baecke PA Questionnaire at ARIC visit 5 (2011-2013) were included. Using the PA component of the American Heart Association’s Life’s Simple 7, participants’ PA levels were categorized as ideal, intermediate, or poor. Cognitive tests were administered at ARIC visits 5 (2011-13), 6 (2016-17), 7 (2018-19), 8 (2020), and 9 (2021-22), and were used to calculate factor scores for global cognitive function and the domains of memory, language, and executive function. Linear mixed models estimated the association of PA with change in cognitive function. Dementia cases were defined by adjudicated review, telephone interviews, informant interviews, hospitalization records, and death certificates. Dementia risk was estimated using Cox proportional hazards regression models. All models adjusted for age, sex, race-center, education, APOE-ε4 allele, smoking, hypertension, body mass index, and diabetes. Results: Participants with ideal or intermediate PA, compared to poor PA, had slower rates of decline in global cognition, but no significant differences in rates of decline across individual cognitive domains. Over a median follow-up of 7.8 years, participants with ideal (adjusted hazard ratio [HR] = 0.62, 95% CI: 0.60, 0.67) or intermediate (adjusted HR = 0.87, 95% CI: 0.80, 0.91) PA had a reduced risk of dementia compared to participants with poor PA. Conclusions: Higher PA levels in late life demonstrated a protective association with long-term global cognitive decline and incident dementia. Although further research with more frequent and objective measures of PA is needed to determine the duration, frequency, and intensity of PA recommended to promote brain health in older adults, our findings suggest that promoting PA among aging populations offers a non-invasive and accessible approach to support brain health.

Introduction: Vascular risk factors of hypertension, diabetes, and smoking in midlife (45-64 years) and possibly early late-life (65-74 years) are associated with increased risk of dementia. Current estimates of population attributable risk for dementia assume vascular factors are independent and do not interact to increase risk of dementia, which may underestimate the combined impact of vascular risk reduction on dementia incidence. In this study, we evaluated the joint association and interaction of vascular risk factors on risk of dementia. Methods: We performed a prospective cohort analysis using the Atherosclerosis Risk in Communities Neurocognitive Study using up to 33 years of follow-up (1987-2020). We used age as baseline (45-54 years, n=7,731; 55-64 years, n=12,273; and 65-74 years, n=6,660). We defined hypertension as systolic blood pressure ≥130 mm Hg, diastolic blood pressure ≥80 mm Hg, or anti-hypertension medication use. We defined diabetes as fasting glucose level ≥126 mg/dL, non-fasting glucose ≥200 mg/dL, self-reported physician diagnosis, or use of any diabetes medication. Current smoking was self-reported. We investigated additive and multiplicative interactions for combinations of vascular risk factors on incident (adjudicated) dementia by age 80 using joint effect variables in Cox regression models stratified by age of risk factor measurement (45-54 years, 55-64 years, or 65-74 years). For interaction analyses, parameter values >1 indicate positive interaction (synergy) and values <1 indicate negative interaction (antagonism). Results: In participants with risk factors measured in these baseline age strata, there were 801 (10%), 995 (8%), and 422 (6%) cases of incident dementia by age 80, respectively. Hypertension and diabetes exhibited positive interactions on the additive and multiplicative scales in midlife (Figure 1). Diabetes and smoking in midlife had positive additive interaction (value >1, Figure 2). All vascular risk factor combinations in late-life had negative interactions on the multiplicative scale (values <1, Figures 1-3). Conclusions: Hypertension with diabetes and diabetes with smoking, all measured in midlife, interact to increase risk of dementia by age 80. The overall impact of reducing midlife vascular risk factors on incident dementia could be sizeable. These findings highlight the importance of multifactor interventions addressing combinations of vascular risk factors for dementia prevention strategies.