Introduction High-quality social interaction is one of the key factors of health and well-being across the lifespan (Hawkley and Cacioppo, 2010; Umberson and Montez, 2010). In late life, supportive friendships can reduce loneliness and protect against cognitive decline (Fratiglioni et al., 2000), while loss of social ties increases vulnerability to isolation and poor outcomes (Holt-Lunstad et al., 2010). Beyond psychological benefits, friendships may promote brain health by buffering stress, supporting healthier lifestyles, and reducing vascular and inflammatory risk factors. These pathways suggest that friendship quality could play a role in neurobiological aging. To investigate this potential link, the present study examined associations between friendship quality and neuroimaging-based brain age measures of gray matter and white matter. Because friendship patterns and brain aging trajectories differ by sex—with women tending to maintain larger, more supportive networks (Shin et al., 2023; Dunbar et al., 2024) and brain aging showing sex-specific trajectories (Scheinost et al., 2015; Wu et al., 2024)—we tested the interaction of sex and friendship quality on gray matter brain age (GMBA) and white matter brain age (WMBA). Methods The sample consisted of 230 older adults (127 females, 103 males) from the Human Connectome Project – Aging (HCP-A) cohort (Bookheimer et al., 2019), aged 65–100 years (Mean = 76.0, SD = 7.8). 66.8% of participants completed 4 years of college or more. Perceived friendship quality and social connectedness was assessed using NIH Toolbox Friendship Survey (Adult v2.0) (Theta score range: –2.52 to 1.65, Mean = 0.10, SD = 0.86). Multilayer Perceptron (MLP)–based brain age models (Chen et al., 2020) were pretrained on independent open imaging datasets [i.e., CamCAN (Taylor et al, 2017) and MCSA (Schwarz et al, 2024), n = 2,418, age = 66.5, SD=14.6)] to estimate GMBA and WMBA. Specifically, GMBA was derived from intracranial volume–adjusted gray matter regional volumes across 56 regions of interest (ROIs) extracted from T1-weighted MRI with the LPBA40 atlas using CAT12 toolbox (Gaser et al., 2024), while WMBA was derived from diffusion tensor imaging (DTI) features (i.e., fractional anisotropy, FA) (Yeh, 2025) across 54 white matter tract bundles based on an HCP1065 Population-Averaged Tractography Atlas (Yeh, 2022). All brain image features were harmonized using the ComBat method (Orlhac, 2022) before model training and inference. Brain age gap is calculated as the difference between the predicted brain age and chronological age (GMBA_gap or WMBA_gap). A positive brain age gap suggests the predicted brain age is older than the chronological age, and a negative brain age gap suggests the predicted brain age is younger than the chronological age. The brain age measures were adjusted to mitigate an age-related bias (Smith et al, 2019). Results There were no significant differences between females and males in age, F(1, 228) = 0.98, p = .323, or education, F(1, 228) = 1.60, p = .220. However, females reported significantly higher friendship quality compared to males, F(1, 228) = 10.61, p = .001. Univariate analyses of variance (ANOVA) were conducted to examine the effects of sex and friendship quality on brain age gap measures (WMBA_gap and GMBA_gap), controlling for education. For WMBA_gap, there was a significant main effect of sex, F(1, 225) = 8.66, p = .004, and a significant sex × friendship interaction, F(1, 225) = 7.99, p = .005. Post hoc analyses stratified by sex showed that higher friendship quality was associated with lower WMBA_gap in females (r = –.18, p = .038), indicating a younger white matter brain age, whereas in males it was associated with higher WMBA_gap (r = .20, p = .049), indicating an older white matter brain age. For GMBA_gap, no significant main effects or interactions were observed (p’s > .42). These findings highlight sex-specific associations between friendship quality and white matter brain aging. Conclusions This study provides novel evidence that friendship quality is differentially associated with white matter, but not gray matter, brain aging in older adults. Higher friendship quality was linked to younger white matter brain age in women, whereas in men it related to older white matter brain age. In women, high-quality friendships may provide greater emotional support, whereas in men, friendships may more often center on shared activities or social comparison, which could influence their impact on stress and resilience. The specificity to white matter is consistent with its vulnerability to vascular and stress-related processes. Because our measure captured perceived friendship quality, these findings reflect subjective evaluations that may not align with structural aspects of social networks. This distinction is important for clinical applications, as subjective perceptions of friendship may highlight psychosocial processes most relevant for intervention. A limitation of this study is that we examined sex rather than gender, which may not fully address the influence of gendered social roles. Overall, friendship quality appears to be a sex-specific factor in white matter aging and a potentially modifiable target for promoting brain health in late life.

Introduction Generalized Anxiety Disorder (GAD) has been associated with disrupted large-scale brain network function, particularly within the executive, emotional, and sensory systems. Functional neuroimaging has uncovered abnormal coupling within and between key neural networks, such as the default mode (DMN), salience, and central executive networks, alongside dysregulated communication between limbic structures and prefrontal regions critical for emotion regulation and cognitive control. Although chronic anxiety affects individuals across their lifespan, the interaction between age-related neural changes and anxiety symptoms remains underexplored, especially in direct comparisons between developmental and geriatric cohorts. This study uses resting-state functional connectivity analyses to investigate neural differences related to GAD in transitional age youth (TAY) and older adults, enabling examination of broad network-level patterns and specific regional interactions. Methods Resting-state functional magnetic resonance imaging (MRI) data from two age cohorts were analyzed. The TAY group consisted of 79 participants (aged 16–26), including individuals with GAD (n=25) and age-matched healthy controls (n=54); depression was treated as a comorbidity. The older adult cohort included 84 participants with an average age of 71.5 years, comprising of individuals with no comorbidities (healthy control, n=24) or multiple overlapping comorbidities: minor depressive state (n=31), major depressive state (n=40), subjective cognitive decline (n=10), or mild cognitive impairment (n=50). For this analysis, anxiety was included as a correlate for each subject grouping in the older adult cohort to enable comparison with the TAY cohort; sex was controlled for in both cohorts. Preprocessing, denoising, and functional analysis of the data was conducted using a combination of in-house scripts built in AFNI, FSL, Python, and SPM, along with the CONN toolbox (22.v2407); the Harvard-Oxford Cortical and Subcortical Structural atlases and the Automated Anatomical Labeling atlas 3 (AAL3) were utilized. Group differences within each cohort, interaction effects of age and GAD, and additive effects of age and GAD on the functional connectivity of 131 ROIs for the younger cohort, as well as a select number of seeds for seed-to-voxel analysis were examined. The TAY cohort was collected in the University of Pittsburgh’s Biomedical Science Tower 3 7T MRI scanner, while the older cohort was collected in two different 3T MRI scanner sites: Ajou University Hospital and Suwon Community Geriatric Mental Health Center. Results There were no significant group differences or interaction effects between age and GAD in the younger cohort; however, a significant additive effect was observed, indicating that age and GAD independently contributed to functional connectivity alterations. Key findings included hypoconnectivity between the left hippocampus and right subgenual anterior cingulate cortex (sACC), and between the right hippocampus and left middle frontal gyrus. Additionally, hyperconnectivity was observed between the posterior cingulate cortex (PCC) and precuneus. See Figures 1–3 and Table 1 for additional connections related to the additive effects of age and GAD. In contrast, the older adult cohort showed several significant hyperconnectivity patterns. These included increased connectivity between: the ventral tegmental area (VTA) and, both, the left amygdala and left inferior temporal gyrus, the precuneus and right inferior temporal gyrus, the PCC and bilateral superior temporal gyri, and the right inferior orbital frontal gyrus and left hippocampus. Conclusions These findings highlight age-dependent alterations in functional connectivity associated with GAD, particularly involving the mesolimbic and mesocortical pathways. In TAY, GAD-related hypoconnectivity between key emotion-regulatory regions—such as the hippocampus, sACC, and middle frontal gyrus—may reflect underdeveloped or dysregulated mesocorticolimbic circuitry, impairing emotion regulation, and top-down cognitive control. Concurrent hyperconnectivity within DMN hubs (PCC–precuneus) suggests heightened internal rumination, a hallmark of anxiety. In older adults, chronic anxiety was associated with widespread hyperconnectivity within mesolimbic and associative networks, including the VTA, hippocampus, amygdala, and inferior temporal cortex. Enhanced coupling between the PCC and bilateral superior temporal gyri, and between the precuneus and inferior temporal regions, points to age-related shifts in DMN integration with affective and associative systems. Notably, opposing hippocampal connectivity patterns—hypoconnectivity with the sACC in TAY versus hyperconnectivity with the inferior orbital frontal gyrus in older adults—may reflect either compensatory reorganization or degenerative changes with aging. Overall, these results suggest that anxiety disrupts emotion and cognitive control circuits differently across the lifespan. While younger individuals exhibit reduced prefrontal–limbic integration, older adults rely more heavily on hyperconnected mesolimbic-affective circuits, potentially reflecting age-related compensation or vulnerability in dopaminergic pathways. Continued research using these cohorts, as well as a mid-life age cohort, may further elucidate the lifespan trajectory of anxiety-related brain network alterations.

Introduction Loneliness is increasingly recognized as a serious and pervasive psychosocial risk factor in late life and has been linked to heightened risks of cognitive decline, depression, and mortality (Hawkley and Cacioppo, 2022; Oken et al., 2024). In late-life brain health, loneliness may be particularly toxic, exerting physiological effects through heightened stress reactivity, systemic inflammation, and vascular dysfunction (Finley 2022; Xia et al. 2018). One neurological substrate of aging is the accumulation of white matter hyperintensities (WMH), brain lesions that act as indicators of cerebral small vessel diseases (Griffanti et al., 2021), and have been linked to cognitive impairments, ​mood disturbances, and​ decreased quality of life. While various psychosocial factors across the lifespan have been implicated in WMH burden, their relative contributions remain unclear in the geriatric population (Taylor et al., 2018). By focusing on loneliness, this study investigates its association with WMH volume to inform therapeutic strategies that target social well-being as a pathway to neurological health in aging populations. Methods We analyzed baseline data from older adults (n=98: median age = 72.4years, IQR, 69.4-76.3; 51% male, 49% female; Table 1), recruited from an ancillary cohort of the parent Arterial Stiffness, Cognition, and Equol trial. The sample was predominantly White (90%), with 10% African American. Educational levels were distributed into three groups: some college or less (22%), completed college (25%), and postgraduate (53%). Participants completed a comprehensive neuropsychological battery, and we employed the UCLA Loneliness Scale (UCLA-LS) as an index of loneliness level. WMH volume was quantified on the T2w FLAIR image using a 2D U-Net machine learning method developed in our lab (Li et al., 2023), normalized by intracranial volume, and log-transformed (nWMH). Linear regression models were used to examine associations between UCLA-LS and WMH burden. Model 1 included unadjusted associations, while Model 2 adjusted for age, sex, race, education, hypertension and diabetes status. Results In univariate analyses (Model 1, Table 2), higher loneliness was marginally associated with greater WMH burden (β = ​0.02, 95% CI [​0.00,0.04], p = 0.060). In fully adjusted analyses (Model 2, Table 2), higher loneliness was significantly associated with greater WMH burden (β = 0.02, 95% CI [0.00, 0.05], p = 0.030), indicating that loneliness was uniquely associated with greater WMH burden, independent of demographic and health characteristics. Sex, race, education, hypertension, and diabetes were not significant predictors, although males showed a marginal trend toward lower WMH burden compared with females (β = -0.39, 95% CI [–0.83, 0.06], p = 0.088). Conclusions In our analysis, loneliness was significantly associated with greater WMH burden, highlighting the potential importance of social and emotional well-being in brain aging. Although other measures were not significant, these results should be interpreted cautiously, given limited power. These results suggest that loneliness may represent a meaningful intervention target aiming at promoting brain health in aging populations. Given the cross-sectional design, causality cannot be inferred; future longitudinal and interventional studies are needed to determine whether reducing loneliness can mitigate WMH progression.

BackgroundPlasma biomarkers for Alzheimer's disease (AD) studies, but much remains unknown about the associations of plasma and PET biomarkers of amyloid-β (Aβ).ObjectiveTo determine the associations of plasma Aβ with PET Aβ accumulation and progression from PET A- to A + .MethodsWe evaluated PET A- participants with baseline plasma Aβ measurements and longitudinal indices of PET Aβ. Linear mixed effects models characterized the association of plasma biomarker outcomes with changes in PET Aβ values. Survival analysis evaluated the ability of baseline plasma biomarker A+/- status to differentiate trajectories of progression from PET A- to A + .ResultsLinear mixed effects models showed significant interactions between time and plasma Aβ42/40 with respect to longitudinal measures of PET Aβ burden. Survival analysis found that A status determined from plasma Aβ42/40 predicted distinct patterns of progression from PET A- to A + .ConclusionsThese models suggest that, in PET A- participants, baseline plasma Aβ42/40 can be used to predict both how much Aβ will accumulate over time and likelihood of becoming PET A + .

Blood-based biomarkers have expanded access to biologically supported diagnosis of Alzheimer's disease (AD), particularly through measurement of amyloid-beta (Aβ) and phosphorylated tau species1-3. Among these, plasma tau phosphorylated at threonine 217 (p-tau217) is currently the leading biomarker recommended by clinical guidelines4-6. However, circulating p-tau217 originates from both central nervous system (CNS) and peripheral tissues7, potentially limiting specificity, particularly in individuals with common age-related comorbidities8. Here we report a next-generation biomarker, brain-derived p-tau217%, which quantifies the proportion of circulating tau that is CNS-derived and phosphorylated at threonine 217. Across neuropathologically defined, Aβ- and tau-neuroimaging-characterized, and memory clinic cohorts, brain-derived p-tau217% consistently identified AD pathology and clinical AD with larger effect sizes, higher discriminative accuracy, and improved sensitivity and specificity, outperforming conventional non-CNS-selective plasma p-tau217, p-tau217/Aβ1-42 and p-tau217% alternatives as well as brain-derived-p-tau217 alone. Furthermore, the CNS-selective biomarker demonstrated more robust prediction of future clinical progression in individuals followed for up to two decades. Importantly, diagnostic performance remained high in older adults with diabetes and cardiovascular disease, populations in which standard p-tau217 showed reduced specificity. Moreover, superiority extended to comparisons against multiple CNS disease-related proteins in targeted proteomic analyses. These findings establish plasma brain-derived p-tau217% as a biologically grounded and clinically robust biomarker that advances molecular definition, detection, and prognosis of Alzheimer's disease.

Cerebrovascular remodeling driven by subtle molecular changes starts early in the asymptomatic stage of Alzheimer's disease (AD). Despite progress in human vascular imaging and postmortem tissue analysis, there is limited data on the early features of small vessel reorganization, particularly in the context of cell-specific molecular drivers. This is largely because of the invasive nature of the tools for direct cellular observation and analysis. Since early detection is key, histopathology falls short with end-point data from people that died in late stages of the disease. This is a critical knowledge gap, because the early vascular processes are thought to be strongly correlated with health outcomes, tipping the scales from mild cognitive impairment to AD. To meet these translational challenges, we performed near life-span in vivo two-photon imaging and MRI of the cerebrovascular tree in a mouse model of amyloidosis. We identified precisely when subtle abnormalities in vessel tortuosity and red blood cell velocity first emerge in the context of differential amyloid accumulation in vessels walls and tissues. We then isolated the brain vessels for transcriptional analysis at this flagship timepoint and performed cross-species analysis linking changes in vascular cells to genes and pathways common to both mice and humans. Importantly, using 7T MRI of aging humans, we directly associated vascular remodeling trajectories of mice and humans and identified a remarkably analogous tortuosity course in the smallest brain vessels. Our integrated framework across scales and species advances neuroimaging biomarker understanding and uncovers early mechanistic routs of dysfunctional angiogenesis and actin-mediated contractility.

Introduction Alzheimer’s disease and related dementias (ADRD) represent an urgent clinical challenge with no curative treatments, making prevention a top priority. Risk is shaped by both non-modifiable factors, such as age, and modifiable factors, including cardiovascular health and lifestyle behaviors. Evidence consistently links these modifiable factors with dementia (Livingston et al., 2024), yet it remains unclear when across the lifespan they exert their greatest influence. Midlife cardiovascular risk has been strongly implicated in later cognitive decline (Gottesman et al., 2014, JAMA Neurology; Joyce et al., 2024, J Hypertens), but the role of cardiovascular health versus social and lifestyle factors (e.g., education, engagement in activities) in late life is less certain. Cross-sectional studies are further complicated by confounding and survival bias, as individuals who reach advanced age represent a healthier subset of the population. To explore whether analytic refinements can yield clearer insights, we applied a propensity score (PS) matching strategy within age-defined strata, balancing observed covariates to reduce confounding and approximate longitudinal inference. This study aimed to test whether a PS framework within age strata could provide a more refined perspective on life-stage–specific risk factors, not as a definitive solution, but as a step toward addressing the methodological limitations inherent in aging research. Methods We analyzed data from 636 participants aged 36–100 years (Mean = 60.5, SD = 15.7) in the Human Connectome Project – Aging (HCP-A) (Bookheimer et al., 2019). Participants were divided into 5-year bins, and three bins were selected to represent different life stages:35–40 years (n = 66; young adulthood),55-60 years (n = 79; midlife), and 80-85 years (n = 57; late-life). Each bin was used once as a reference group, and separate propensity score analyses were performed for each reference. For each reference bin, propensity scores were estimated with logistic regression including demographic covariates (sex, race, ethnic group, testing site, handedness score) and blood markers (alanine aminotransferase, aspartate aminotransferase, bilirubin, insulin, chloride, potassium, creatinine, vitamin D). Nearest-neighbor matching was applied to create groups comparable on demographic and blood markers, leaving age as the main difference. Within each matched dataset (35–40: n₁=122; 55–60: n₂=140; 80–85: n₃=104), linear regressions were conducted to test associations of two outcomes: (1) cognition, measured with the Montreal Cognitive Assessment (MoCA), and (2) hippocampal volume, derived from FreeSurfer-processed T1-weighted images and normalized for intracranial volume, with predictors including age, education, body mass index (BMI), low density lipoprotein (LDL), blood pressure, and exercise. For comparison, two regressions were also conducted in the unmatched full cohort (n = 636), one for MoCA and one for hippocampal volume, using the same predictors. Results Distinct life-stage patterns emerged based on the choice of propensity matching reference. With the 35–40 year reference group, higher MoCA scores were linked with lower BMI (β=-0.219, p=0.027) and higher education (β=0.228, p=0.011), and greater hippocampal volume was associated with female sex (β=-0.291, p < 0.001). With the 55–60 year reference group, hippocampal volume was strongly associated with lower LDL (β=-0.153, p=0.040) and female sex (β=-0.207, p=0.005), while higher MoCA scores were related to higher education (β=0.143 p=0.085). With the 80–85 year reference group, hippocampal volume was marginally associated with lower blood pressure (β=-0.155, p=0.057) and female sex (β=-0.124, p=0.088), and MoCA scores with higher education (β=0.236, p=0.011). Unmatched analyses suggested broader associations (e.g., exercise with cognition, LDL with hippocampal volume) that attenuated after PS adjustment, underscoring the importance of balancing covariates. Conclusions This exploratory study suggests that modifiable risk factors for cognition and hippocampal volume vary by life stage: cardiovascular factors (BMI, LDL, blood pressure) showed stronger effects when referencing to midlife, whereas education emerged as more influential with a late-life reference (80–85 years). These shifting patterns highlight the clinical importance of a life-stage perspective in appreciating dementia prevention. identifying vascular health factors using a midlife reference and cognitive reserve with a later life match life-stage models. While limited by sample size and exploratory design, our findings reinforce that risk factors are not static, and timing matters for both scientific investigation and patient counseling.

Introduction White matter hyperintensities (WMH) are a key MRI biomarker in aging and neurodegeneration, including Alzheimer’s disease (AD), where higher WMH burden relates to worse cognition and clinical progression. Accurate assessment of the cerebral vasculature is therefore critical for diagnosing and understanding neurological disorders, especially those with subtle vascular contributions as seen in AD and the broader VCID spectrum. Beyond global vascular risk, vessel diameter and tortuosity has been linked to cerebrovascular disease phenotypes and to WMH burden, motivating morphology-based analyses within the brain’s arterial tree. In this study, we quantify bilateral ACA/MCA/PCA diameter and tortuosity on TOF-MRA and test their associations with WMH burden. Methods This study included 49 older adults who underwent MRI on a 7T Siemens Terra-like scanner at the University of Pittsburgh (N=49, 33F/16M, mean age at baseline (SD) = 71.3 (8.2) years, age range [54–96] years). All images were acquired on a 7 T Siemens scanner. Time-of-flight MRA (TOF-MRA) used 0.38 × 0.38 × 0.38 mm³ voxels (354 slices; ∼12 min), and same-session 3D T2-FLAIR provided input for WMH segmentation. WMH masks were generated fully automatically with wmh_seg, a Transformer-based U-Net previously developed by our lab; total WMH volume was computed from the final masks, normalized by intracranial volume (ICV), and log-transformed to yield nWMH. The whole-brain arterial tree was segmented from TOF-MRA using VesselMapper, yielding a centerline graph with per-point radius; diameters were computed as 2 times radius times voxel size and summarized along graph segments. eICAB provided Circle-of-Willis (CoW) labels (L/R MCA, L/R PCA, ACA). To capture proximal–distal structure, we designated the first n BFS hops from each CoW seed as large (proximal) vessels (Figure left panel) and the remainder as small (distal) vessels (Figure middle panel). For each branch, tortuosity was computed using the distance metric (path length / chord length) and diameter as above; within-subject features were then formed as the mean diameter and mean tortuosity for each category (L/R MCA large/small, L/R PCA large/small, ACA large/small). In addition, we derived overall large-vessel (pooled across all territories’ large segments), overall small-vessel (pooled across all small segments), and overall (pooled across the entire arterial tree) mean diameter and mean tortuosity. Results Adjusting for age and sex, greater overall vessel tortuosity was associated with increased WMH burden (β = 2.14, p = 0.0138). This association was primarily driven by small vessel tortuosity, as reflected in the overall mean small vessel tortuosity measure (β = 0.809, p = 0.0218). In particular, left MCA small vessel tortuosity showed the strongest relationship with WMH, with higher tortuosity strongly associated with greater lesion burden (β = 1.55, p = 0.0019). By contrast, right MCA big vessel tortuosity demonstrated an inverse relationship with WMH (β = –0.604, p = 0.0457), suggesting that increased tortuosity of larger right MCA vessels was linked to fewer lesions. Conclusions In this study, we show that intracranial vascular geometry tracks WMH burden. Overall tortuosity—and especially small-vessel tortuosity—was associated with greater lesions, led by left MCA small vessels, whereas right MCA big-vessel tortuosity related inversely. These territory- and scale-specific patterns suggest arterial morphology encodes cerebrovascular injury risk and may provide a noninvasive biomarker and mechanistic target for WMH in aging.

Psychosocial factors may shape brain health long before clinical symptoms of age-related cognitive decline appear. This theoretically grounded narrative review selectively synthesizes findings on psychosocial influences on white matter hyperintensities (WMH), which are neuroimaging markers of cerebrovascular aging, in otherwise healthy adults. We selectively synthesized literature on psychosocial factors, including psychological stress, depression, anxiety, trauma, and neuroticism, and their associations with WMH burden in healthy midlife and older adult samples. Evidence links multiple psychosocial factors to WMH burden, though findings are inconsistent across studies. Integrating these findings, we propose a conceptual framework in which chronic psychological stress contributes to WMH development and progression through hypertension-mediated mechanisms. The framework situates this pathway within the broader context of biological aging, cumulative stress exposure, and social determinants of health, including race and socioeconomic status. This framework is offered to guide future research on the psychosocial determinants of cerebrovascular aging and advance understanding of inequalities in brain and vascular health.

Growing evidence suggests vascular dysfunction plays a critical role in the early stages of Alzheimer's disease, commonly associated with amyloid-β deposition. This vascular dysfunction is particularly relevant in the context of cerebral amyloid angiopathy, where amyloid-β accumulates within cerebral vessel walls. Notably, sex differences impact progression of both Alzheimer's disease and cerebrovascular dysfunction, with post-menopausal females displaying increased small vessel disease burden and diminished carbon dioxide reactivity compared to older males and pre-menopausal females. Moreover, the cerebrovasculature is a target of sex hormones where they exert influence in numerous vascular functions and pathologies across lifespan. Combined, cerebrovascular dysfunction along with amyloid-β deposition may have differential effects on sex. Despite observational studies in humans, preclinical mechanistic and functional research on sex-specific vascular differences in Alzheimer's disease has been limited. In this near-lifespan longitudinal study, we investigated age and sex-specific neurovascular coupling and carbon dioxide reactivity in a transgenic mouse model expressing chimeric mouse/human amyloid precursor and mutant human presenilin 1 (APP/PS1) and control mice using widefield optical imaging. Neurovascular coupling was probed via whisker stimulation and then vascular reactivity was measured using hypercapnic challenge. During whisker stimulation, neuronal activity was measured through GCaMP6f fluorescence change, while vascular response was quantified via haemoglobin-based optical intrinsic signal. Carbon dioxide reactivity was evaluated by measuring dilatory changes of vessel diameters across the cerebrovascular tree. In vivo two-photon microscopy was used to longitudinally measure cerebral amyloid angiopathy vessel coverage and amyloid-β tissue plaque volume. We observed that APP/PS1 mice exhibited attenuated neurovascular coupling during whisker stimulation and this response worsened through lifespan compared to controls. Compared to controls, APP/PS1 mice exhibited decreased carbon dioxide reactivity with age. No sex differences between control mice were observed in the neurovascular response to whisker, whereas during hypercapnia, control females had higher carbon dioxide reactivity than control males. While both APP/PS1 males and females showed reduced dilatory responses with age, APP/PS1 females exhibited this decrease in small arteries, whereas APP/PS1 males experienced decreased dilation in larger arteries. Diminished vascular reactivity in APP/PS1 mice was associated with increased cerebral amyloid angiopathy and amyloid-plaque burden. This study highlights sex-specific pathophysiology's of vascular dysfunction across the lifespan. Our findings underscore needs to incorporate sex differences in preclinical Alzheimer's disease research, given the rising importance of vascular contributions to cognitive impairment and dementia. Our findings have important implications for developing targeted, age and sex-specific biomarkers and therapeutics for cerebrovascular health in Alzheimer's disease.

INTRODUCTION:Given the predominance of imaging and plasma biomarkers in Alzheimer's disease observational studies and clinical trials, it is critical to understand the differences between these biomarkers across racialized groups. METHODS:A total of 260 older adults without dementia racialized as Black and/or African American (AA) and non-Hispanic white (NHW), ranging in age from 50 to 90 years (68.8 ± 9.1 years), were evaluated for differences in plasma amyloid-β (Aβ) 42/Aβ40, p-tau181, p-tau217, p-tau231, neurofilament light chain (NfL), and glial fibrillary acidic protein (GFAP) as well as Aβ positron emission tomography (PET) and magnetic resonance (MR) imaging-derived cortical thickness using Mann-Whitney U tests and analysis of covariance (ANCOVA). RESULTS:Both Mann-Whitney tests and ANCOVA found significant differences between groups racialized as AA or NWH with respect to global 11[C]-Pittsburgh Compound B (PiB) standardized uptake value ratio (SUVR), cortical thickness values, p-tau181, and p-tau231 values (p < 0.05). DISCUSSION:Racialization should be given more consideration in AD clinical research, particularly when biomarker results are used for inclusion or exclusion criteria for clinical trials and qualification in clinical practice. HIGHLIGHTS:Global 11[C]-Pittsburgh compound B (PiB) standardized uptake value ratio (SUVR), cortical thickness, p-tau181, and p-tau231 differed between groups Differences were unaffected by age, sex, apolipoprotein E *4 (APOE*4), education, and Mini-Mental State Examination (MMSE) score Racialization needs more consideration in Alzheimer's disease clinical research Additional work is needed to understand the sources of biomarker differences.

INTRODUCTION:Late-life cognitive impairment and depression frequently co-occur and share many symptoms. However, the specific neural and clinical factors contributing to both their common and distinct profiles in older adults remain unclear. METHODS:We investigated resting-state correlates of cognitive and depressive symptoms in older adults (n = 248 and n = 95) using clinical, blood, and neuroimaging data. We computed a connectivity matrix across default mode, executive control, and salience networks. Cross-validated elastic net regression identified features reflecting cognitive function and depressive symptoms. These features were validated on a held-out dataset. RESULTS:We discovered that white matter hyperintensities and nine overlapping nodes spanning all three networks are associated with both cognitive function and depressive symptoms, including left amygdala, left hippocampus, and bilateral ventral tegmental area. DISCUSSION:Our findings reveal intertwined neural nodes influencing cognitive impairment and depressive symptoms in late life, offering insights into shared characteristics and potential therapeutic targets. HIGHLIGHTS:Resting-state neuroimaging markers are associated with symptoms of cognitive decline and late-life depression. Symptom-associated connectivity alterations were present across three major brain networks of interest, including the salience, default mode, and executive control networks. Some regions of interest are associated with both cognitive function and depressive symptoms, including the left amygdala, left hippocampus, and bilateral ventral tegmental area.

Amyloid-PET imaging tracks the accumulation of amyloid beta (Aβ) deposits in the brain. Amyloid plaques accumulation may begin 10 to 20 years before the individual experiences clinical symptoms associated with Alzheimer's diseases (ad). Recent large-scale genome-wide association studies reported common risk factors associated with brain amyloidosis, suggesting that this endophenotype is driven by genetic variants. However, these loci pinpoint to large genomic regions and the functional variants remain to be identified. To identify new risk factors associated with brain amyloid deposition, we performed whole-genome sequencing on a large cohort of European descent individuals with amyloid PET imaging data (n = 1,888). Gene-based analysis for coding variants was performed using SKAT-O for amyloid PET as a quantitative endophenotype that identified genome-wide significant association for APOE (P = 2.45 × 10-10), and 26 new candidate genes with suggestive significance association (P < 5. 0 × 10-03) including SCN7A (P = 7.31 × 10-05), SH3GL1 (P = 7.56 × 10-04), and MFSD12 (P = 8.51 × 10-04). Enrichment analysis highlighted the lipid binding pathways as associated with Aβ deposition in brain driven by PITPNM3 (P = 4.27 × 10-03), APOE (P = 2.45 × 10-10), AP2A2 (P = 1.06 × 10-03), and SH3GL1 (P = 7.56 × 10-04). Overall, our data strongly support a connection between lipid metabolism and the deposition of Aβ in the brain. Our study illuminates promising avenues for therapeutic interventions targeting lipid metabolism to address brain amyloidosis.

Background:Tau accumulation in Alzheimer's disease is associated with short term clinical progression and faster rates of cognitive decline in individuals with high amyloid-β deposition. Defining an optimal threshold of tau accumulation predictive of cognitive decline remains a challenge. Objective:We tested the ability of regional tau PET sensitivity and specificity thresholds to predict longitudinal cognitive decline. We also tested the predictive performance of thresholds in the proposed new NIA-AA biological staging for Alzheimer's disease where multiple levels of tau positivity are used to stage participants. Methods:18F-flortaucipir scans from 301 non-demented participants were processed and sampled. Four cognitive measures were assessed longitudinally. Regional standardized uptake value ratios were split into infra- and suprathreshold groups at baseline using previously derived thresholds. Survival analysis, log rank testing, and Generalized Estimation Equations assessed the relationship between the application of regional sensitivity/specificity thresholds and change in cognitive measures as well as tau threshold performance in predicting cognitive decline within the new NIA-AA biological staging. Results:The meta temporal region was best for predicting risk of short-term cognitive decline in suprathreshold, as compared to infrathreshold participants. When applying multiple levels of tau positivity, each subsequent level of tau identified cognitive decline at earlier timepoints. Conclusions:When using 18F-flortaucipir, meta temporal suprathreshold classification was associated with increased risk of cognitive decline, suggesting that abnormal tau deposition in the cortex predicts decline. Likewise, the application of multiple levels of tau clearly predicts the distinctive cognitive trajectories in the new NIA-AA biological staging framework.

Importance:Emerging evidence suggests that severe acute respiratory syndrome, COVID-19, negatively impacts brain health, with clinical magnetic resonance imaging (MRI) showing a wide range of neurologic manifestations but no consistent pattern. Compared with 3 Tesla (3T) MRI, 7 Tesla (7T) MRI can detect more subtle injuries, including hippocampal subfield volume differences and additional standard biomarkers such as white matter lesions. 7T MRI could help with the interpretation of the various persistent post-acute and distal onset sequelae of COVID-19 infection. Objective:To investigate the differences in white matter hyperintensity (WMH), hippocampal subfields volumes, and cognition between patients hospitalized with COVID-19 and non-hospitalized participants in a multi-site/multi-national cohort. Design:Original investigation of patients hospitalized with COVID-19 between 5/2020 and 10/2022 in 3 USA and 1 UK medical centers with follow-up at hospital discharge. Participants:A total of 179 participants without a history of dementia completed cognitive, mood and other assessments and MRI scans. Exposure:COVID-19 severity, as measured by hospitalization vs no hospitalization. Main Outcomes and Measures:7T MRI scans were acquired. All WMH and hippocampal subfield volumes were corrected for intracranial volumes to account for subject variability. Cognition was assessed using a comprehensive battery of tests. Pearson correlations and unpaired t-tests were performed to assess correlations and differences between hospitalized and non-hospitalized groups. Results:We found similar WMH volume (4112 vs 3144mm³, p=0.2131), smaller hippocampal volume (11856 vs 12227mm³, p=0.0497) and lower cognitive and memory performance, especially the MoCA score (24.9 vs 26.4 pts, p=0.0084), duration completing trail making test B (97.6 vs 79.4 seconds, p=0.0285), Craft immediate recall (12.6 vs 16.4 pts, p<0.0001), Craft delay recall (12.0 vs 15.6 pts, p=0.0001), and Benson figure copy (15.2 vs 16.1 pts, p=0.0078) in 52 patients hospitalized for COVID-19 (19[37%] female; mean[SD] age, 61.1[7.4] years) compared with 111 age-matched non-hospitalized participants (66[59%] female; mean[SD] age, 61.5[8.4] years). Conclusions and Relevance:Our results indicate that hospitalized COVID-19 cases show lower hippocampal volume when compared to non-hospitalized participants. We also show that WMH and hippocampal volumes correlate with worse cognitive scores in hospitalized patients compared with non-hospitalized participants, potentially indicating recent lesions and atrophy. Key Points:Question: Do white matter hyperintensity burden, hippocampal whole and subfield volumes, and cognition differ between patients hospitalized with COVID-19 versus participants without hospitalization?Findings: We found no significant difference in white matter hyperintensity volume, but hippocampal volume was reduced, and cognitive and memory performance were worse in those hospitalized for COVID-19 compared with age-matched non-hospitalized group (either mild COVID-19 or no COVID-19 reported). In the hospitalized group, increased white matter hyperintensity and reduced hippocampal volumes are significantly higher correlated with worse cognitive and memory scores.Meaning: Adults hospitalized for COVID-19 had lower hippocampal volumes and worse cognitive performance than adults with COVID-19 that did not lead to hospitalization or without reported COVID-19 infection.

White matter hyperintensity (WMH) lesions on brain MRI images are surrogate markers of cerebral small vessel disease (CSVD). Longitudinal studies examining the association between diabetes and WMH progression have yielded mixed results. Thus, in this study we investigated the association between HbA1c, a biomarker for the presence and severity of hyperglycemia, and longitudinal WMH change after adjusting for known risk factors for WMH progression. We recruited 64 participants from South Korean memory clinics to undergo brain MRI at the baseline and a two-year follow-up. We found: First, higher HbA1c was associated with greater global WMH volume (WMHV) changes after adjusting for known risk factors (B = 7.7E-04, p = 0.025); Second, the association between baseline WMHV and WMHV progression was only significant at diabetic levels of HbA1c (p < 0.05, when HbA1c > 6.51%), and non-APOE ɛ4 carriers showed a stronger association between HbA1c and WMHV progression (B = -2.59E-03, p = 0.004); Third, associations of WMHV progression with HbA1c were particularly apparent for deep WMHV change (B = 7.17E-04, p < 0.01) compared to periventricular WMHV change, and for frontal (B = 5.00E-04, p < 0.001) and parietal (B = 1.534-04, p < 0.05) WMHV change compared to occipital and temporal WMHV change. In conclusion, higher HbA1c levels were associated with greater two-year WMHV progression, especially in non-APOE ɛ4 participants or those with diabetic levels of HbA1c. These findings demonstrate that diabetes may potentially exacerbate cerebrovascular and white matter disease.