Baseline Cognition Research Articles

Association of MRI markers of vascular injury and neurodegeneration and cognitive decline in diverse cohorts: Findings from KHANDLE and STAR

AbstractBackgroundMRI biomarkers of vascular injury and neurodegeneration are associated with cognitive decline, but relatively little is known about cognitive change in diverse populations. Our objective is to examine associations of hippocampal volume and white matter integrity on cognitive change over 3 years.MethodKHANDLE (age 65+) and STAR (age 50+) are harmonized cohorts of racially/ethnically diverse long‐term members of Kaiser Permanente Northern California. A random sub‐sample of participants received neuroimaging with 3T MRI to measure regional brain volumes and white matter integrity. Hippocampal volume (cm3), log white matter hyperintensities (log‐WMH) (cm3), free water (FW) (cm3), and fractional anisotropy (FA) MRI measures were normalized on intracranial volume and age at scan. Cognitive decline was assessed across three waves using the Spanish English Neuropsychological Assessment Scale measuring executive function (EF) and verbal episodic memory (VEM). Linear mixed models allowing for random intercept and slope estimated the associations of MRI metrics and cognition adjusting for baseline age, education, gender/sex, interview mode (in‐person vs phone), and interactions of these covariates with time from baseline, to adjust for effects of covariates on both baseline cognition and cognitive change.ResultAmong 592 participants with neuroimaging, 59% were women, 47% were college educated, and 52% were Black adults (Table 1). Hippocampal volume was positively associated with baseline VEM (β = 0.268; CI = 0.153, 0.383) and EF change (β = 0.051; CI = 0.009, 0.094) (Table 2, Plot 1). FW was negatively associated with baseline EF (β = ‐6.561; CI = ‐10.758, ‐2.364), baseline VEM (β = ‐5.557; CI = ‐10.03, ‐1.084) and EF decline (β = ‐1.667; CI = ‐3.268, ‐0.066). FA was positively associated with baseline EF (β = 6.222; CI = 3.217, 9.228) and baseline VEM (β = 6.944; CI = 3.768, 10.119). Log‐WMH was negatively associated with baseline EF (β = ‐0.068; CI = ‐0.116, ‐0.020) and VEM (β = ‐0.082; CI = ‐0.132, ‐0.032).ConclusionAll MRI markers were associated with baseline cognition, but only hippocampal volume and FW were associated with cognitive change. MRI measures of neurodegeneration and white matter integrity may be predictive of cognition and short‐term decline in diverse groups of older adults. Future analyses will examine associations with cognitive impairment.

Persistent financial adversity across adulthood and cognitive ageing: A lifecourse investigation.

AbstractBackgroundCognitive decline in older age has significant health, economic and social consequences and there is evidence that financial adversity might play a role. Financial adversity, for many, is not an acute but a chronic stressor. However, little is known about the effect of persistence exposure to financial adversity across the lifecourse on cognitive ageing.MethodUsing data from the 1946 British birth cohort (N = 2,759) with an embedded neuroimaging study (Insight46; N = 356‐468), we examined the prospective association between financial adversity (low household income, financial hardships; 26‐53 years) and cognitive performance (69 years) and decline (53‐69 years), as well as volumetric measures of brain health (69‐71 years) in older adulthood. Cognitive decline was modelled using latent growth curves, with derived intercepts and slopes used in subsequent regression analyses, adjusted for sex, childhood socioeconomic circumstances (SEC), baseline cognition and internalising symptoms, and educational attainment. Sex, childhood SEC and APOE‐ɛ4 genetic risk were further tested as moderators in regression models using interaction terms. This study has been pre‐registered (https://doi.org/10.17605/OSF.IO/6Y9NV).ResultIncreased exposure to financial adversity showed dose‐response associations with slower verbal memory decline and lower processing speed and verbal memory at 69 years (Table 1). Each increased period of exposure to low household income or financial hardships was associated with a 0.11 and 0.08 standard deviation (SD) decrease in processing speed, and a 0.32 and 0.22 SD decrease in verbal memory at 69 years, respectively. Males persistently exposed to financial hardships showed slower processing speed decline but increased brain atrophy compared to females. Those whose fathers were in unskilled profession or unemployed in childhood, and who also persistently experienced low income themselves showed greater brain atrophy (Fig. 1). APOE‐ɛ4 carriers who persistently experienced financial adversity were also at greater risk for brain atrophy (Fig. 2).ConclusionPersistent financial adversity is directly associated with cognitive performance and may be differentially associated with brain atrophy depending on demographic and genetic factors. Reducing the frequency of financial adversity over the lifecourse may improve cognitive performance and protect against genetic susceptibility to brain atrophy and subsequent dementia risk in the ageing population.

Differences in the association between midlife hypertension status and late‐life cognition and cognitive change: results from the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) Study

AbstractBackgroundThis study evaluates the association of midlife hypertension status and domain‐specific late‐life cognitive change allowing for possible differences by sex.MethodThe Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) cohort is a multi‐ethnic study of 1,712 Kaiser Permanente Northern California members age 65+ at enrollment in 2017. For 1,696 participants, we categorized the last examined clinical blood pressure measures from 1964‐1985 (mean age = 37.5) as hypotensive, normotensive, prehypertensive, or hypertensive. Verbal episodic memory and executive function were assessed using Spanish and English Neuropsychological Assessment Scales across three interview waves over 4 years and scores were z‐standardized to the baseline. Sex‐stratified linear mixed models with random intercepts estimated the associations of midlife hypertension status with domain‐specific baseline cognition and cognitive change using years since baseline as timescale. Models adjusted for age at blood pressure measurement, KHANDLE baseline age, race, ethnicity, education, practice effects, and interview mode (phone vs in‐person).ResultA greater proportion of men than women had midlife pre‐hypertension (44% vs 31%) or hypertension (21% vs 11%) (Table 1). Among men and women, midlife hypertension status was not associated with baseline executive function or verbal episodic memory (Table 2). Compared to men with normal midlife blood pressure, those with midlife hypotension (b = ‐0.10, 95% CI: ‐0.17, ‐0.04) or midlife hypertension experienced steeper declines in executive functioning (b = ‐0.05, 95% CI: ‐0.10, ‐0.004). Among women, only midlife hypertension was associated with decline in executive function (b = ‐0.04 95% CI: ‐0.09, 0.003), though not significantly different from normotensive women. Compared to normotensive men, those with midlife hypotension (b = ‐0.17, 95%CI: ‐0.27, ‐0.07) and those with pre‐hypertension (b = ‐0.06, 95% CI: ‐0.12, ‐0.004) experienced steeper decline in verbal episodic memory. Among women, only midlife hypertension was associated with steeper declines executive function (b = ‐0.04, 95% CI: ‐0.09, 0.003), though not significantly different from normotensive women. Though effect estimates of midlife hypertension status on cognitive outcomes differed by sex, these differences were not statistically significant.ConclusionBoth hypotension and hypertension 30‐50 years prior to cognitive assessments predicted accelerated cognitive decline in older men. Patterns were less consistent for women but not statistically distinguishable from associations among men.

The effect of blood pressure on domain‐specific cognitive changes in older Black Americans: findings from the Baltimore Study of Aging – Patterns of Cognitive Aging

AbstractBackgroundPrevious research suggests that hypertension shape cognitive health outcomes in older Blacks (Whitfield 2008). However, few studies have explored the relation between objective and subjective blood pressure (BP) measures and cognitive decline. This study examined whether systolic, diastolic, and self‐reported BP were associated with changes in cognition.MethodData are drawn from the Baltimore Study of Black Aging—Patterns of Cognitive Aging. At baseline, 602 Black participants (48‐95 years) were enrolled. At follow‐up, approximately 3 years later, 450 participants were re‐interviewed. Cognition included 5 domains: working memory, processing speed, verbal memory, vocabulary, and inductive reasoning. Using an oscillometer automated device (A&D model UA‐ 767), three readings of orthostatic BP were collected and mean systolic (SBP; range = 82‐230.3 mmHg) and diastolic BP (DBP; range = 51‐136.7 mmHg) values were calculated for each participant. BP was treated as a continuous variable. Participants were also asked “Have you ever been told by a doctor or nurse that you have high blood pressure? (yes/no response)” Data were analyzed using multiple linear regression, controlling for sociodemographic and health conditions.ResultMajority of the sample was female (25.4% male) and had at least a high school education (57.1%). Systolic BP was inversely associated with inductive reasoning at follow‐up (b = ‐0.208, p = 0.023; see Table 1), but was not related to any other cognitive domains, after adjusting for age, sex, education, depressive symptoms, comorbidities, diastolic BP, and baseline cognition. Diastolic BP and self‐reported hypertensive status were both unrelated to changes in each of the five cognitive domains (all models, p> 0.05).ConclusionThe present study adds to our understanding of cognitive aging among older Black Americans. Results support previous research, pointing to an independent effect of BP on domain‐specific cognitive changes (Buford 2016), but also extend the literature by highlighting a more pronounced negative effect of systolic BP. These findings suggest that the pathway to cognitive dysfunction is contributed to by BP. Several key mechanisms – including increased risk for stroke and “vascular dementia” (age‐ and hypertension‐related cerebral vasculature changes) – may explain the influence of hypertension on cognitive decline that warrant further study in Black Americans.

Neurofilament Light and mean Free Water as prognostic enrichment biomarkers for cognitive worsening and small vessel disease progression

AbstractBackgroundPlasma neurofilament light (NfL) and free water (FW), a DTI‐derived measure, are considered, individually, sensitive markers of axonal damage that relate to cognition. A significant remaining challenge for clinical trials resides in identifying biomarkers able to detect, simultaneously, early SVD abnormalities and baseline cognitive impairment, particularly in non‐demented participants, who are also at risk of cognitive worsening and SVD progression during the trial.MethodThis study includes 200 non‐demented individuals from the MarkVCID‐UH3 cohort who had baseline NfL and baseline and follow‐up (1.3±0.5 years) cognitive assessment and MRI (Table 1). We used a composite measure of global cognitive function (GCF) using scores from the NACC‐UDS. WMH volumes were log‐transformed and corrected for intracranial volume. Annual change (Δ) in GCF and WMH was computed using baseline and follow‐up measures. A threshold, determined from 3 independent cohorts, was applied to FW measures to group individuals (low: <0.22 and high > = 0.22). Low and high NfL levels were defined using age‐based NfL z‐scores (low←0.5SD and high> = ‐0.5SD). We finally grouped individuals with both high NfL and FW levels (MULTIBIO+) vs. individuals with at least one biomarker with low level (MULTIBIO‐). We used linear regressions with either baseline GCF, baseline WMH, ΔGCF or ΔWMH as the dependent variable and, separately, baseline FW, NfL and MULTIBIO as the independent variable. Analyses were adjusted for relevant confounding variables.ResultsWe found that, as compared with MULTIBIO‐, MULTIBIO+ group had significantly lower baseline GCF and higher baseline WMH (p = 0.033 and p<0.001), as well as accelerated ΔGCF and larger ΔWMH (pvalues<0.01, Table 2, Figure 1). Individually, although FW was associated with baseline WMH and accelerated ΔGCF, neither FW nor NfL levels simultaneously explained baseline cognition and SVD burden, or simultaneous ΔGCF and ΔWMH (Table 2).ConclusionThese findings emphasize the potential of combining NfL and FW measures as a prognostic enrichment biomarker panel for identifying individuals with both early SVD abnormalities and cognitive impairment and who are also at risk of cognitive worsening and substantial SVD burden progression during clinical trials.Data collection and sharing for this project was funded by NINDS/NIA as part of the MarkVCID Consortium (U24NS100591, UH2NS100599, UH2NS100605, UH2NS100588, UH2NS100608, UH2NS100606, UH2NS100598, UH2NS100614)

Social health and subsequent cognitive capability: examining the potential mediating roles of depression symptoms and inflammatory biomarkers

AbstractBackgroundSocial health (SH) markers, including marital status, contact frequency, network size, and social support, have been linked with increased cognitive capability. However, the underlying mechanisms remain poorly understood. We aim to investigate whether depression symptoms and inflammatory biomarkers mediate associations between SH and cognitive outcomes.MethodWe used data from waves 1‐9 of the English Longitudinal Study of Ageing, involving 7,136 participants aged 50 or older at baseline. First, we examined associations between SH (wave 1) and depression and inflammatory biomarkers (C‐reactive protein (CRP) and fibrinogen) (wave 2) using linear regression models. Second, we tested associations between a) SH and b) depression and inflammation with subsequent standardised verbal fluency and memory in wave 3 and change between waves 3‐9, indexed using slopes derived from multilevel models. We adjusted for age, sex, socio‐economic position, cardiovascular disease, basic and instrumental activities of daily living, health behaviours, and baseline depression symptoms and cognition. We will also conduct causal mediation analysis.ResultAll SH markers, except contact frequency, were associated with lower subsequent depression, but not inflammatory biomarkers. Greater contact frequency (e.g. once‐twice a week vs <once per year: β = 0.18 [0.01, 0.36]) and less negative support (β = 0.02 [0.00, 0.03]) were associated with higher verbal fluency. Larger network size (>6 people vs none: β = 0.007SD/year [0.001, 0.012]), less negative (β = 0.001SD/year [0.001, 0.002]) and more positive support (β = 0.001SD/year [0.000, 0.001]) were linked with slower memory decline, and more positive support predicted slower verbal fluency decline (β = 0.001SD/year [0.000, 0.001]). Depression symptoms were associated with lower memory and verbal fluency, and faster memory decline (β = ‐0.001SD/year [‐0.001, ‐0.000]) and verbal fluency (β = ‐0.001SD/year [‐0.001, ‐0.000]). CRP was associated with lower verbal fluency (β = ‐0.02 [‐0.04, 0.00]), whereas fibrinogen was linked with faster memory decline (β = ‐0.001SD/year [‐0.003, ‐0.000]).ConclusionDepression symptoms and SH showed associations with subsequent cognitive capability and change. SH was linked with lower depression, but not inflammatory biomarkers. Findings highlight the potential for depression to underpin associations between SH and cognition, a pathway which we will test using causal mediation analysis. We will also examine whether findings replicate in the Swedish National Study of Aging and Care in Kungsholmen.

FC40: Social health and subsequent cognitive capability: examining the potential mediating roles of depression symptoms and inflammatory biomarker

Objective:Social health (SH) markers, including marital status, contact frequency, network size, and social support, have been linked with increased cognitive capability. However, the underlying mechanisms remain poorly understood. We aim to investigate whether depression symptoms and inflammatory biomarkers mediate associations between SH and cognitive outcomes.Methods:We used data from waves 1-9 of the English Longitudinal Study of Ageing, involving 7,136 participants aged 50 or older at baseline. First, we examined associations between SH (wave 1) and depression and inflammatory biomarkers (C-reactive protein (CRP) and fibrinogen) (wave 2) using linear regression models. Second, we tested associations between a) SH and b) depression and inflammation with subsequent standardised verbal fluency and memory in wave 3 and change between waves 3-9, indexed using slopes derived from multilevel models. We adjusted for age, sex, socio-economic position, cardiovascular disease, basic and instrumental activities of daily living, health behaviours, and baseline depression symptoms and cognition. We will also conduct causal mediation analysis.Results:All SH markers, except contact frequency, were associated with lower subsequent depression, but not inflammatory biomarkers. Greater contact frequency (e.g. once-twice a week vs <once per year: β=0.18 [0.01, 0.36]) and less negative support (β=0.02 [0.00, 0.03]) were associated with higher verbal fluency. Larger network size (>6 people vs none: β=0.007SD/year [0.001, 0.012]), less negative (β=0.001SD/year [0.001, 0.002]) and more positive support (β=0.001SD/year [0.000, 0.001]) were linked with slower memory decline, and more positive support predicted slower verbal fluency decline (β=0.001SD/year [0.000, 0.001]). Depression symptoms were associated with lower memory and verbal fluency, and faster memory decline (β=-0.001SD/year [-0.001, -0.000]) and verbal fluency (β=-0.001SD/year [-0.001, -0.000]). CRP was associated with lower verbal fluency (β=-0.02 [-0.04, 0.00]), whereas fibrinogen was linked with faster memory decline (β=-0.001SD/year [-0.003, -0.000]).Conclusion:Depression symptoms and SH showed associations with subsequent cognitive capability and change. SH was linked with lower depression, but not inflammatory biomarkers. Findings highlight the potential for depression to underpin associations between SH and cognition, a pathway which we will test using causal mediation analysis. We will also examine whether findings replicate in the Swedish National Study of Aging and Care in Kungsholmen.

The associations between glial biomarkers and cognitive changes in individuals at risk of Alzheimer’s disease

AbstractBackgroundIn preclinical Alzheimer’s disease (AD), previous evidence suggests that glial reactivity may affect cognitive performance. This study aimed to test whether glial markers cerebrospinal fluid (CSF) GFAP, YKL‐40, sTREM2, and plasma GFAP are associated with cognitive change in cognitively unimpaired individuals at risk of AD. We hypothesized that higher levels of glial markers may be associated with better cognitive outcomes.MethodWe included 353 individuals from the ALFA+ cohort with available baseline (BL) glial biomarkers and cognition at follow‐up (FU) 3 years later (Table 1). CSF Aβ42/40, pTau181, YKL‐40, GFAP, and sTREM2, were analyzed using the exploratory Roche NeuroToolKit, a panel of automated robust prototype immunoassays. Plasma GFAP was determined with the Simoa HD‐X (Quanterix) platform using the commercial Neurology 4‐Plex E assay (N4PE). Amyloid pathology was studied with Centiloids derived from [18F]flutemetamol PET. Rate of change ([FU‐BL]/time) in the Preclinical Alzheimer’s Cognitive Composite (PACC) and domain‐specific cognitive composites was used as the dependent variable, glial biomarkers at BL as predictors, and age, sex and education as covariates in linear models. Interaction terms between each glial biomarker and different AD pathology were also explored. Aβ+ was defined as CSF Aβ42/40 <0.071.ResultHigher baseline CSF sTREM2 was associated with a positive PACC rate of change (p = 0.002) (Table 2). These results remained significant after controlling for CSF Aβ42/40, pTau181 and Centiloid levels (all p < 0.01). Moreover, CSF sTREM2 at baseline was associated with a better memory outcome (p = 0.012), independently of amyloid and tau (all p < 0.02), and better executive functioning (p = 0.046), independently of tau pathology (p = 0.011). These positive associations remained in the Aβ‐ group regardless of AD pathology, but not in the Aβ+ group, in which some negative associations were observed when adjusting for Centiloids (YKL‐40, p = 0.024) (Figure 1).ConclusionHigher TREM2‐mediated microglial response in individuals at risk of AD is associated with better cognitive performance. These associations were driven by the Aβ‐ group. These results may have important implications for the design of clinical interventions targeting TREM2 in this early stage of AD.

Association of plasma Alzheimer’s and neurodegenerative disease biomarkers with brain volume loss and cognitive decline

AbstractBackgroundAlthough plasma Alzheimer’s disease (AD) and neurodegenerative disease biomarkers have been associated with reduced cognition and lower brain volume, few studies have examined whether plasma biomarkers measured in predominately cognitively unimpaired adults are associated with longitudinal brain volume loss and cognitive decline. Using data from the Baltimore Longitudinal Study of Aging (BLSA), we examined whether plasma biomarkers of AD pathology (Aß42/40, phosphorylated tau [pTau‐181]), reactive astrogliosis (glial fibrillary acidic protein [GFAP]), and neuronal injury (neurofilament light chain [NfL]) were associated with longitudinal brain volume loss and cognitive decline.MethodPlasma biomarkers (Aß42/40, pTau‐181, GFAP, NfL) were measured using Quanterix SIMOA assays. Baseline visits were considered the earliest visits that had concurrent 3T MRI scans and plasma biomarker measurements, and the earliest visits that had concurrent cognitive assessments and plasma biomarker measurements. Composite Z scores were computed to reflect 5 cognitive domains (Figure 1). Linear mixed effects models adjusted for baseline age, sex, race, education, estimated glomerular filtration rate, and total intracranial volume (brain volume analyses), were used to examine the association of plasma biomarkers with baseline and longitudinal brain volume and cognition. False Discovery Rate correction was used to adjust for multiple comparisons.ResultBrain volume analyses included 628 participants (mean age±SD: 71.0±10.2; Table 1) with an average of 3.3 MRI scans over 4.7 years. Cognition analyses included 686 participants (mean age±SD: 71.4±10.1) with an average of 3.9 cognitive assessments over 5.7 years. Although Aß42/40 was unrelated to brain volume changes, higher pTau‐181 was associated with longitudinal declines in total gray matter volume and regional declines in medial temporal regions. Higher GFAP was associated with longitudinal increases in ventricular volume, whereas NfL was not associated with brain volume change (Table 2). Both lower Aß42/40 and higher pTau‐181 were associated with a decline in verbal memory. Additionally, lower Aß42/40 was associated with a decline in visuospatial performance. Further, higher GFAP was associated with a decline in verbal fluency (Figure 1).ConclusionPlasma biomarkers of AD pathology and astrogliosis may serve as markers of future brain atrophy and cognitive decline.

Synergistic effects of age at menopause and vascular risk on cognitive decline in postmenopausal women

AbstractBackgroundMenopause‐related estrogen depletion is linked to greater vascular risk, Alzheimer’s disease (AD) pathology, and dementia risk in women. Age‐at‐menopause and vascular risk may together influence the likelihood of developing AD in women. We investigated whether earlier menopause and elevated vascular risk independently or synergistically contribute to cognitive decline in postmenopausal women compared to men.MethodWe used data from the Canadian Longitudinal Study on Aging. We included 7,678 postmenopausal women (mean age = 64.6±8.49 years, mean age‐at‐menopause = 50.1±5.02 years) and 7,678 age‐matched men (mean age = 64.6±8.49 years). Sex was operationalized using self‐reported sex at birth. Vascular risk was calculated using a summary score tallying the presence/absence of hypertension, diabetes, smoking, obesity, cardiovascular disease, and stroke. Participants completed cognitive assessments at baseline and 3‐year follow‐up. Self‐reported age‐at‐menopause was categorized into tertiles: early (<49), average (49‐52), and late (>52). Linear models tested independent and interactive effects of age‐at‐menopause and vascular risk on 3‐year cognition, adjusting for baseline cognition, baseline age, education, APOE4, and test language (English/French).ResultIn women, vascular risk and age‐at‐menopause were independently associated with cognitive decline (vascular risk: β = ‐0.107, p<.001). Relative to average age‐at‐menopause, early menopause was associated with greater cognitive decline (β = ‐0.026, p = .03), whereas late menopause was associated with less decline (β = 0.025, p = .04). In men, vascular risk was also associated with greater cognitive decline (β = ‐0.112, p<.001). Next, we examined synergistic effects of age‐at‐menopause and vascular risk in postmenopausal women. Early (vs. average) menopause and elevated vascular risk interacted to accelerate cognitive decline (β = ‐0.224, p = .002). There was no interaction between late (vs. average) menopause and vascular risk (β = ‐0.073, p = .3). In a model comparing sexes (Figure 1), vascular risk was more strongly associated with cognitive decline in women with early menopause compared to men (β = ‐0.127, p = .03). In women with average or late menopause, associations of vascular risk with cognitive decline did not significantly differ from men (average: β = ‐0.010, p = .07; late: β = ‐0.024, p = .7). Sensitivity analyses additionally adjusting for physical activity, race/ethnicity, depression, alcohol, and hormone therapy yielded similar results.ConclusionEarly menopause and vascular processes may synergistically contribute to increased AD risk in women. These findings have implications for sex‐specific dementia prevention strategies.

Predictors of hospital mortality in older adults with and without dementia: a necessary step towards personalized prognostication

AbstractBackgroundDementia has been associated with increased risk of hospital mortality (1). However, predictors of hospital mortality for patients with dementia have not been widely addressed. We aimed to investigate predictors of hospital mortality according to preexisting clinical dementia in older adults.MethodsWe analyzed data from a prospective cohort study in Sao Paulo, Brazil. We included acutely ill patients aged ≥60 years admitted to a tertiary university hospital between 2009 and 2020. Candidates underwent comprehensive geriatric assessments in the first 24 hours of hospitalization. Preexisting clinical dementia (after this, referred to as dementia) was determined by a Clinical Dementia Rating (CDR) ≥1 (2). We investigated independent predictors of hospital mortality according to dementia using multivariable Poisson regressions. We further analyzed our multivariable models' accuracy in predicting hospital mortality using area under the ROC curves (AUC).ResultsWe included 3,863 participants (mean age = 80; female = 62%). Overall, 1,761 participants (46%) had dementia. Mortality was 11% in participants without dementia and 21% in those with dementia (p<0.001). In participants without dementia, mortality was independently predicted by age, multimorbidity, functional status, delirium, nutritional status, and serum albumin (Table 1). However, in participants with dementia, mortality was also predicted by illness acuity and creatinine. Finally, we observed that our multivariable model had an AUC = 0.88 (95%CI = 0.86‐0.90) in participants without dementia and 0.80 (0.78‐0.83) in those with dementia (Figure 1).ConclusionOlder adults with dementia were twice as likely to die in the hospital than those without dementia. Moreover, independent predictors of hospital mortality were not necessarily shared between patients with and without dementia. Given the importance of prognostic assessments for decision‐making processes, researchers should consider exploring new algorithms that differ according to baseline cognition.

A longitudinal study of polygenic score and cognitive function decline considering baseline cognitive function, lifestyle behaviors, and diabetes among middle-aged and older US adults

BackgroundGenomic study of cognition decline while considering baseline cognition and lifestyle behaviors is scarce. We aimed to evaluate the impact of a polygenic score for general cognition on cognition decline rate, while considering baseline cognition and lifestyle behaviors, among the general population and people with diabetes, a patient group commonly affected by cognition impairment.MethodsWe tested associations of the polygenic score for general cognition with annual changing rates of cognition measures in 8 years of follow-up among 12,090 White and 3100 Black participants of the Health and Retirement Study (HRS), a nationally representative sample of adults aged 50 years and older in the USA. Cognition measures including word recall, mental status, and total cognitive score were measured biannually. To maximize sample size and length of follow-up, we treated the 2010 wave of survey as baseline, and follow-up data until 2018 were analyzed. Baseline lifestyle behaviors, APOE status, and measured cognition were sequentially adjusted. Given racial differences in polygenic score, all analyses were conducted by race.ResultsThe polygenic score was significantly associated with annual changing rates of all cognition measures independent of lifestyle behaviors and APOE status. Together with age and sex, the polygenic score explained 29.9%, 15.9%, and 26.5% variances of annual changing rates of word recall, mental status, and total cognitive scores among Whites and explained 17.2%, 13.9%, and 18.7% variance of the three traits among Blacks. Among both White and Black participants, those in the top quartile of polygenic score had the three cognition measures increased annually, while those in the bottom quartile had the three cognition measures decreased annually. After further adjusting for the average cognition assessed in 3 visits around baseline, the polygenic score was still positively associated with annual changing rates of all cognition measures for White (P ≤ 2.89E − 19) but not for Black (P ≥ 0.07) participants. In addition, among participants with diabetes, physical activity offset the genetic susceptibility to decline of mental status (interaction P ≤ 0.01) and total cognitive scores (interaction P = 0.03).ConclusionsPolygenic score predicted cognition changes in addition to measured cognition. Physical activity offset genetic risk for cognition decline among diabetes patients.

Biomarkers for the prediction of cognitive decline in atrial fibrillation patients

Abstract Background Patients with atrial fibrillation (AF) are at increased risk of cognitive decline. Biomarkers might help to identify the highest risk patients. Purpose To investigate associations of a broad panel of biomarkers with cognitive decline in a large population of AF patients. Methods We enrolled 1440 AF patients with available baseline biomarkers and cognitive testing by the Montreal Cognitive assessment (MoCA) score at inclusion and at ≥2 yearly follow-ups over three years. Cognition over time was assessed by modelling of a slope over three years. Cognitive decline was defined as a decreasing slope based on 1 standard deviation of the baseline MoCa score. We investigated the associations of biomarkers and age with cognitive decline in univariate logistic regression models by AUCs. We used Lasso modelling to build combined prediction models with biomarkers, clinical variables and both. Results Mean age was 72 years, 75% were male, 47% had paroxysmal AF and 90% were anticoagulated. During a follow-up of 3 years, cognitive decline occurred in 93 patients (6.5%). Patients with cognitive decline were older (77 vs 72 years, p=0.001), more often had paroxysmal AF (50 vs 47%, p=0.007) or a history of stroke (24 vs 11%, p=0.001), but there were no differences in anticoagulation (94 vs 90%, p=0.289). The three biomarkers with the highest univariate AUC for cognitive decline were Growth Differentiation Factor (GDF)-15 (0.67 [0.62-0.72]), Cystatin C (0.67 [0.61-0.72]) and high-sensitivity Troponin T (0.65 [0.60-0.70]), while age had the highest overall AUC (0.70 [0.64-0.75]) (Figure). The combined prediction model with the highest AUC of 0.69 (0.67-0.72) included GDF-15, age and baseline cognition. Conclusion Over 3 years, 6.5% of AF patients had cognitive decline despite a high rate of anticoagulation. Besides age, GDF-15, Cystatin C and high-sensitivity Troponin T had the highest predictive value for cognitive decline.Univariate AUCs for biomarker/ageFigure legend

Cognition in older adults in Uganda: Correlates, trends over time and association with mortality in prospective population study.

Dementia is an important and growing issue in sub-Saharan Africa, but epidemiological data are lacking. Risk factors may differ from other regions due to high stroke incidence and HIV prevalence. Understanding the epidemiology of cognition in older adults in Africa is crucial for informing public health strategies to improve the lives of people with dementia and their carers. The Wellbeing of Older People Study in Uganda is an open cohort of adults aged 50+ with and without HIV, established in 2009. Detailed socio-demographic and health data have been collected at four waves spanning 10 years, including cognitive assessment using internationally validated WHO-recommended tests: verbal recall, digit span, and verbal fluency. Mortality data was collected until the end of the fourth wave (2019). We examined associations of low baseline cognition scores and changes in cognition score over time using random effects modelling, care needs of people with lower cognition scores, and the relationship between cognition score and mortality. Data were collected on 811 participants. Older age, lower educational attainment, lower socio-economic position, and extremes of BMI were associated with lower cognition scores. Cognition scores declined faster at older ages, but rate of decline was not associated with cardiovascular disease or HIV at baseline. People with lower cognition scores required more assistance with Activities of Daily Living, but mortality rates were similar across the range of cognition scores. The crucial next step will be to investigate types and presentation of clinical dementia in this cohort, so we can better understand the clinical relevance of these findings to inform public health planning.

What is the association between adverse childhood experiences and late-life cognitive decline? Study of Healthy Aging in African Americans (STAR) cohort study

ObjectivesAdverse childhood experiences (ACEs) are associated with higher risk of chronic disease, but little is known about the association with late life cognitive decline. We examined the longitudinal association between...

17 Emotional and Instrumental Support as Protective Factors in Cognitive Aging Among Black and Hispanic/Latinx Older Adults

Objective:Social support may protect against Alzheimer’s disease and related dementias (ADRD), potentially through emotional or instrumental support elements. Black and Hispanic/Latinx older adults bear a disproportionate burden of ADRD. However, independent effects of emotional and instrumental support on cognition, a primary indicator of ADRD risk, are largely understudied in these groups. Guided by the differential vulnerability hypothesis – the theoretical framework which posits that systemic racism disadvantages Black and Hispanic/Latinx individuals’ health – we hypothesize that emotional and instrumental support may be particularly important to protect against worse cognition for Black and Hispanic/Latinx older adults, who often have fewer resources due to these inequalities (e.g., wealth, educational opportunities) to otherwise maintain health. Using the NIH Toolbox Emotion Module measures of emotional (e.g., the extent to which individuals can rely on others in challenging times) and instrumental support (e.g., the extent to which individuals can rely on others for assistance in daily activities), we aimed to identify positive social support factors (i.e., emotional and instrumental support) that may protect against ADRD risk (i.e., longitudinal executive function and memory performance) among Black and Hispanic/Latinx older adults.Participants and Methods:Participants were 362 Black and 265 Hispanic/Latinx adults aged 65-89 (63% female, average age=75) from the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) Study who completed baseline and up to two additional waves of assessments (every 1.5 years), including questionnaires, neuropsychological evaluations, and the NIH toolbox. Predictors included baseline covariates (i.e., age, language of test administration, gender, education, income, self-rated health) and NIH toolbox emotional and instrumental support variables. Outcomes were baseline and longitudinal memory (visual and verbal episodic memory) and executive functioning (verbal fluency and working memory) composites from the Spanish and English Neuropsychological Assessment Scales (SENAS). Latent growth curve models were conducted separately in Black and Hispanic/Latinx participants to estimate effects of emotional and instrumental support on baseline cognition and subsequent change in each domain.Results:Black participants reported greater emotional support. There were no group differences in levels of instrumental support. Greater instrumental support was associated with better initial memory (standardized β= .194, 95%CI: [.063, .325]) among Black participants but not among Hispanic/Latinx participants. In Hispanic/Latinx participants, greater emotional support was associated with better initial executive functioning (standardized β= .215, 95%CI: [.079, .350]. Emotional support was not associated with either cognitive domain in Black participants. There were no associations between emotional or instrumental support on cognitive change in either group.Conclusions:Results point to differences between Black and Hispanic/Latinx older adults in the impact of specific aspects of social support on different cognitive domains. Positive associations between instrumental support and baseline memory in Black participants and between emotional support and executive functioning in Hispanic/Latinx participants suggest unique cognitive consequences of social support across groups. Differences in the role of specific types of social supports may be useful in identifying intervention targets specifically for Black and Hispanic/Latinx older adults, who are disproportionately affected by ADRD. Future research will examine these constructs using multiple group models to test these associations more rigorously.

84 Feasibility and Validity of Remote Digital Assessment of Multi-Day Learning in Cognitively Unimpaired Older Adults

Objective:Unsupervised remote digital cognitive assessment makes frequent testing feasible and allows for measurement of learning across days on participants’ own devices. More rapid detection of diminished learning may provide a potentially valuable metric that is sensitive to cognitive change over short intervals. In this study we examine feasibility and predictive validity of a novel digital assessment that measures learning of the same material over 7 days in older adults.Participants and Methods:The Boston Remote Assessment for Neurocognitive Health (BRANCH) (Papp et al., 2021) is a web-based assessment administered over 7 consecutive days repeating the same stimuli each day to capture multi-day-learning slopes. The assessment includes Face-Name (verbal-visual associative memory), Groceries-Prices (numeric-visual associative memory), and Digits-Signs (speeded processing of numeric-visual associations). Our sample consisted of200 cognitively unimpaired older adults enrolled in ongoing observational studies (mean age=74.5, 63% female, 87% Caucasian, mean education=16.6) who completed the tasks daily, at home, on their own digital devices. Participants had previously completed in-clinic paper-and-pencil tests to compute a Preclinical Alzheimer’s Cognitive Composite (PACC-5). Mixed-effects models controlling for age, sex, and education were used to observe the associations between PACC-5 scores and both initial performance and multi-day learning on the three BRANCH measures.Results:Adherence was high with 96% of participants completing all seven days of consecutive assessment; demographic factors were not associated with differences in adherence. Younger participants had higher Day 1 scores all three measures, and learning slopes on Digit-Sign. Female participants performed better on Face-Name (T=3.35, p<.001) and Groceries-Prices (T=2.00, p=0.04) on Day 1 but no sex differences were seen in learning slopes; there were no sex differences on Digit-Sign. Black participants had lower Day 1 scores on Face-Name (T=-3.34, p=0.003) and Digit Sign (T=3.44, p=0.002), but no racial differences were seen on learning slopes for any measure. Education was not associated with any measure. First day performance on Face-Name (B=0.39, p<.001), but not learning slope B=0.008, p=0.302) was associated with the PACC5. For Groceries-Prices, both Day 1 (B=0.27, p<.001) and learning slope (B=0.02, p=0.03) were associated with PACC-5. The Digit-Sign scores at Day 1 (B=0.31, p<.001) and learning slope (B=0.06, p<.001) were also both associated with PACC-5.Conclusions:Seven days of remote, brief cognitive assessment was feasible in a sample of cognitively unimpaired older adults. Although various demographic factors were associated with initial performance on the tests, multi-day-learning slopes were largely unrelated to demographics, signaling the possibility of its utility in diverse samples. Both initial performance and learning scores on an associative memory and processing speed test were independently related to baseline cognition indicating that these tests’ initial performance and learning metrics are convergent but unique in their contributions. The findings signal the value of measuring differences in learning across days as a means towards sensitively identifying differences in cognitive function before signs of frank impairment are observed. Next steps will involve identifying the optimal way to model multi-day learning on these subtests to evaluate their potential associations with Alzheimer’s disease biomarkers.

The Effects of Midlife Acute and Chronic Stressors on Black-White Differences in Cognitive Decline.

Midlife stressors may be particularly consequential for cognitive performance and disparities in cognitive decline. This study examined Black-White differences in trajectories of cognition among middle-aged adults and the effects of acute and chronic stressors on these trajectories. Data come from 4,011 cognitively healthy individuals aged 51-64 (620 Black and 3,391 White) who participated in the 2006-2018 waves of the Health and Retirement Study. Stressors included a count of recent life events and measures of financial strain and everyday discrimination. Global cognition was assessed using a modified version of the Telephone Interview for Cognitive Status. Linear mixed models with random slopes and intercepts assessed change in cognition over time. Race-by-time, race-by-stressor, time-by-stressor, and race-by-stressor-by-time interactions were assessed as were quadratic terms for time and each stressor. After adjusting for sociodemographic, health behaviors, and health-related factors, Black respondents had lower initial cognitive performance scores (b = -1.75, p < .001) but experienced earlier but slower decline in cognitive performance over time (Black × Time2 interaction: b = 0.02, p < .01). Financial strain, discrimination, and recent life events each had distinct associations with cognitive performance but did not influence racial differences in levels of or change in cognition over time. Middle-aged Black adults have lower initial cognition levels and experience earlier but less accelerated cognitive decline compared to White middle-aged adults. Midlife acute and chronic stressors influence baseline cognition but do so in different ways. Future research should examine the influence of other stressors on racial differences in cognitive decline at other points in the life course.

Neighborhood resources and risk of cognitive decline among a community-dwelling long-term care population in the U.S.

ObjectiveTo examine the associations between neighborhood resources (i.e., number of restaurants, recreation centers, or social services for seniors and persons with disability per land area) and cognitive decline among a community-dwelling long-term care population and whether they differ by baseline cognition status. Study designProspective longitudinal cohort study. MethodsWe used a longitudinal dataset that assessed over a two-year period older adults receiving state-funded home- and community-based services in Michigan Metropolitan areas (N = 9,802) and applied nonlinear mixed models with a random intercept with Poisson distribution. ResultsCognitively intact older adults were less likely to experience cognitive decline when they resided in resource-rich neighborhoods, compared to those cognitively intact but living in neighborhoods that lacked resources. But their cognitively impaired or dementia-diagnosed counterparts did not similarly benefit from living in neighborhoods with rich resources. ConclusionsNeighborhood resources may be an important aspect of intervention to mitigate cognitive decline before older adults become cognitively impaired.

Brain-derived neurotrophic factor as a biomarker in cancer-related cognitive impairment among adolescent and young adult cancer patients

Brain-derived neurotrophic factor (BDNF) improves cognitive function by stimulating neurogenesis and neuroplasticity. We hypothesize that higher plasma BDNF levels are protective against cognitive toxicity among adolescent and young adult cancer patients (15–39 years old). In a prospective, longitudinal study, we recruited 74 newly diagnosed cancer and 118 age-matched non-cancer controls who completed the Cambridge Neuropsychological Test Automated Battery (CANTAB), Functional Assessment of Cancer Therapy-Cognitive Function questionnaire (FACT-Cog) and blood draws. Plasma BDNF was quantified using an enzyme-linked immunosorbent assay. Genomic DNA from buffy coat was genotyped for BDNF Val66Met. Most cancer participants were diagnosed with breast (24%) and head/neck (22%) cancers. After adjusting for sociodemographic variables (age, gender, race, marital status, education years), cancer participants had lower BDNF levels (ng/mL) at baseline (median: 10.7 vs 21.6, p < 0.001) and 6-months post-baseline (median: 8.2 vs 15.3, p = 0.001) compared to non-cancer controls. Through linear mixed modelling adjusted for sociodemographic variables, baseline cognition, fatigue, psychological distress, and time, we observed that among cancer participants, lower baseline BDNF levels were associated with worse attention (p = 0.029), memory (p = 0.018) and self-perceived cognitive abilities (p = 0.020) during cancer treatment. Met/Met was associated with enhanced executive function compared to Val/Val (p = 0.012). Plasma BDNF may serve as a predictive biomarker of cancer-related cognitive impairment.