TARCC 2023 Scientific Symposium

Poster #32

IDENTIFICATION OF GENETIC REGIONS ASSOCIATED WITH ALZHEIMER 'S DISEASE BIOMARKERS BLOOD CONCENTRATIONS IN MEXICAN AMERICAN FAMILIES.

Presenting author: Marcio Augusto Afonso de Almeida, PhD (Research Assistant Professor, University of Texas Rio Grande Valley)

Background: Dementias are heterogeneous age-associated neurodegenerative disorders generally included in the broad term, Alzheimer's disease, and related dementias (ADRD). ADRD affects individuals of all ethnicities, but Hispanic individuals show a 1.5-fold higher risk when compared to non-Hispanic whites. Many AD risk biomarkers have been proposed but not much is known about how genetics elements control their expression. We used a set of 2000 completely sequenced Mexican Americans for the identification of genetic regions associated with AD biomarkers blood concentration. Methods: We applied a mean-based endophenotype ranking value (ERV) method tailored for relatively rare diseases and extended pedigrees. A set of 70 GOBS subjects were diagnosed as ADRD cases (h2= 0.75, p = 2.6*10-5). We quantified the blood concentrations, using a Quanterix elisa assay, of four AD candidate biomarkers AB40, AB42, TAU and NFL. We identified genomic regions associated with those four biomarkers using a linear mixed model as implemented in SOLAR on a set of 28 million SNPs. Candidate SNPs identified were extensively annotated and their individual EQTL contribution to the expression of flanking genes defined. Results: The heritability component of each AD biomarker was defined, and the most significant estimate observed for NFL (h2= 0.33, p=6.96*e-17) where the individual's age had a large effect (2.97*e-126). We identified two genome-wide significant associations for markers rs242557 (3.11*e-14) and rs242562 (1.34*e-13) both SNPs are common and located respectively in the first and second introns of the gene MAPT. Both markers are in a conserved region with chromatin accessibility for transcription factors, suggesting regulatory functions. We tested the association between both SNPS and flanking gene expression and identify strong evidence of cis-regulation with the expression of the antisense-RNA (KANSL1_AS1) of the gene KANSL1. Conclusion: The identification of genetic markers associated with AD biomarkers is key to provide a reliable and cost-effective genetic test for AD risk for the Hispanic population. The gene KANSL1 encodes a subunit of histone acetylation enzyme and has been associated with Koolen-de Vries Syndrome and cognitive impairment. The genetic associations are promising, but demand an independent validation to confirm their impact for the AD risk estimation.

Funding Disclosure: NIH P30AG066546; NIH R21AG077501

Poster #33

FEDEXOSOME: MITOCHONDRIAL DNA & MICRO RNA PROFILING OF NEURONAL-ENRICHED PACKAGES IN ALZHEIMER'S DISEASE

Presenting author: Courtney Hall, PhD (Postdoctoral Fellow, University of North Texas Health Science Center)

Background: Alzheimer's disease (AD) is a complex neurodegenerative disorder characterized by impaired memory, diminished cognitive abilities, and accumulation of amyloid beta plaques and hyperphosphorylated tau tangles in the brain. AD progression is insidious with the earliest clinical symptoms appearing decades after key neuropathological changes have occurred. Accurate diagnosis during the transitional period between disease onset and clinical manifestation could provide critical insight into AD pathogenesis and enable development of effective therapeutic strategies to prevent irreversible neuronal death. Evidence suggests that early alterations in the AD brain (oxidative stress, neuroinflammation, aberrant gene expression) can propagate to local and distal cells through the biological packages secreted by neurons. These neuronal-enriched extracellular vesicles (NEEVs) can cross the blood-brain barrier and are capable of mediating systemic inflammation in response to CNS-sourced stress through their mitochondrial DNA (mtDNA) and microRNA (miRNA) cargo. Circulating NEEVs may therefore serve as an easily accessible, early indicator of the forthcoming neuropathological changes in the AD brain. This study aims to develop a higher throughput workflow for profiling mtDNA and miRNA in plasma NEEVs. Methods: Human plasma samples were processed using a two-step method involving (1) precipitation of total exosomes with ExoQuick (SBI) and (2) immunoprecipitation-based enrichment of NEEVs with a biotinylated antibody against a well-established neuronal surface marker (CD171). Captured NEEVs were stained and analyzed using flow cytometry as well as nanoparticle tracking analysis (NTA). DNA and RNA were then extracted for mtDNA quantification and miRNA sequencing. Result: The protocol modifications implemented in this project were compared based on particle size, distribution, and purity as well as absolute mtDNA load and spike-in normalized miRNA read counts. These data not only demonstrate the feasibility of profiling the mtDNA and miRNA from 500µL of plasma NEEVs but have also resulted in a higher throughput workflow for processing TARCC patient samples. Conclusion: This work represents the first successful attempt to simultaneously quantify mtDNA and sequence miRNA in the relatively small subpopulation of plasma EVs that originate from neurons. Future studies will harness the protocol developed herein to assess the biomarker potential of NEEVs in the TARCC cohort.

Funding Disclosure: TARCC (RS00049)

Poster #34

EFFECTS OF TRANSCRANIAL INFRARED LASER STIMULATION ON BRAIN RHYTHMIC ELECTRICAL ACTIVITY IN YOUNGER AND OLDER ADULTS

Presenting author: Dariella Fernandez, MA (Graduate Student, University of Texas at Austin)

Background: Transcranial Infrared Laser Stimulation (TILS) is a non-invasive intervention that has been found to modulate mitochondrial respiration and cellular functions in brain neurons. In healthy, young adults, eight minutes of TILS has been shown to improve memory and attention. In older adults, TILS is being investigated as a potential intervention against cognitive decline and dementia. The research on the electrophysiological effects of TILS is in the incipient stage and little is known about what electrophysiological effect TILS has on the aging brain. The objectives of the study are to: 1) Examine the acute electrophysiological effects of TILS during and after TILS administration in younger adults and older adults, 2) Investigate the relationship between baseline cognitive function and electrophysiological effects of TILS in older adults, and 3) Investigate if a single 8-minute administration of TILS in older adults can impact cognitive functioning. Methods: This study will employ a sham-controlled, randomized, single blind experimental design. We will recruit a cohort of younger adults (<45 years old) and older adults (>45 years old) who do not have dementia nor severe cardiovascular disease. All participants will complete a 13-minute electroencephalogram (EEG) that measures before, during and after 8-minutes of TILS (250 mW/cm2, 1064 nm) or sham treatment. Older adults will additionally be asked to complete three quick cognitive tasks before TILS treatment, and will repeat the cognitive tasks 3-4 days later. We will calculate both within-group and between-group changes in EEG power and use non-parametric permutation tests to test for significance. Expected Results: We expect that both younger and older adults will show TILS induced, dose-dependent increases in alpha and beta power and that older adults will have a greater benefit from TILS. We also expect that older adults with lower baseline cognitive scores will show a greater benefit from TILS and that a single administration of TILS will improve performance on the cognitive tasks. Conclusion: The results from this study will help us further understand the mechanistic link between photobiomodulation and the cognitive enhancing benefits from TILS, how TILS affects the aging brain, and can help guide future clinical applications of TILS.

Funding Disclosure: Oskar Fischer Project; Elhapa Foundation

Poster #35

HEALTHY AGING IMPACTS THE OSCILLATORY DYNAMICS UNDERLYING VERBAL WORKING MEMORY PERFORMANCE

Presenting author: Lin Guo, BA (Medical Student, UT Southwestern Medical Center)

Background: Normal cognitive aging begins in midlife and is associated with decreases in processing speed, working memory capacity, and attention. In contrast to neurodegenerative disorders including Alzheimer's disease and related dementias, age-related cognitive decline does not progress to impair functioning in activities of daily living. Building upon prior work done on a smaller sample of mostly Caucasian males (Proskovec, Human Brain Mapping, 2016), the present study will characterize the impact of healthy aging on the oscillatory dynamics supporting successful verbal working memory (VWM) performance in a more diverse cohort using magnetoencephalographic (MEG) functional brain imaging. Methods: Fifty-four cognitively normal (CN) adults without subjective cognitive complaints (15 Black, 30 female, 59 mean age, 38-81 age range) performed a modified Sternberg VWM task during MEG. 6 letters were simultaneously presented in a 2x3 grid for 2.0 s (encoding), followed by an empty grid for 3.0 s (maintenance), and a test letter for 0.9 s (retrieval). Subjects responded via button pad whether the test letter had been in the original 6-letter set. MEG data underwent standard pre-processing and co-registration to the subject's anatomic MRI. Artifact-free, correct trials will be transformed into the time-frequency domain to identify significant, sensor-level oscillatory responses, which will then be reconstructed in brain space using a beamformer approach. The resulting whole-brain maps of oscillatory activity will be entered into regression analyses to assess the effects of age, while controlling for race, sex, and education level. Hierarchical regression was also performed to model the effect of age on VWM task performance with race, sex, and education level as covariates. Results: Average reaction time increased with age after adjusting for race, sex, and education level (p=0.003); however, age was not predictive of task accuracy (p>0.05). The oscillatory dynamics underlying this well-preserved VWM task performance in older adults despite age-related decrement in processing speed will be investigated. We expect to observe stronger, more generalized recruitment of predominantly left-lateralized VWM brain areas with advancing age. Conclusion: Understanding neural changes in cognitively healthy, older adults will help delineate pathological alterations and differentiate abnormal from expected cognitive aging during early, clinically indistinguishable stages.

Funding Disclosure: None

Poster #36

THE ROLE OF AGE AND EDUCATION ON NEUROCOGNITIVE FUNCTIONING IN HISPANIC AND NON-HISPANIC INDIVIDUALS IN RURAL WEST TEXAS

Presenting author: Lauren Elliott, BA (Graduate Student, Texas Tech University)

Background: Research indicates education and age influence neurocognitive functioning; however, little is known about education and age affecting Hispanic and non-Hispanic populations, specifically, within underserved rural individuals in West Texas. Methods: 1864 participants (752 non-Hispanic, 1053 Hispanic; Mage = 59.68 years, SDage =12.21) in Project FRONTIER (Facing Rural Obstacles to Healthcare Now Through Intervention, Education, & Research) completed five Neuropsychological tests: The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), Trails Making Test A and B, and Clock Drawing 1 and 2. Age was dichotomized at 65 to assess older (65+) vs younger adults (<65). Education was grouped by completed years of schooling: elementary school (0-5 years), middle school (6-8), high school (9-12), and higher education (13-20). Results: The MANOVA identified significant main effects for age (Roy's Largest Root = 12.52, F(5, 1,363), p < .001, η2p = 0.044) and ethnicity (Roy's Largest Root = 61.29, F(5, 1,363), p < .001, η2p = 0.184) on indicators of neurocognitive functioning. Post-hoc analysis revealed older participants had lower overall neurocognitive functioning on RBANS (M=82.52, SD=16.75) than younger participants (M=85.24, SD=14.95). Post-hoc analysis revealed Hispanic participants had lower overall neurocognitive functioning (M=78.49, SD=14.26) than non-Hispanic participants (M=92.11, SD=15.07). The second MANOVA identified a significant interaction between ethnicity and education on neurocognitive functioning (Roy's Largest Root = 5.11, F(5, 1,548), p < 0.001, η2p = 0.016). Non-Hispanic individuals who reported less education had worse neurocognitive functioning compared to participants with more education. However, Hispanic participants did not have as significant increase between low education compared to higher education. Conclusions: Lower age and higher education were linked to greater neurocognitive functioning. However, educational attainment had a disproportionate lower impact on neurocognitive functioning for Hispanic individuals compared to non-Hispanic individuals. These results are consistent with the theory of Minorities' Diminished Returns, which posits educational attainment has weaker protective effects for ethnic minority groups. It could be hypothesized the education system is structured to provide greater benefits for non-Hispanic individuals, compared to Hispanic individuals. Due to potential systematic biases, education status may only serve as a protective factor for those it was developed: non-Hispanic, white individuals.

Funding Disclosure: None

Poster #37

INTRAINDIVIDUAL VARIABILITY ACROSS NEUROPSYCHOLOGICAL TESTS AND BETWEEN COGNITIVE DOMAINS: WHAT ARE WE REALLY MEASURING?

Presenting author: Bonnie M. Scott, PhD (Assistant Professor, The University of Texas at Austin Dell Medical School)

Background: Intraindividual variability (IIV) in cognitive performance is a non-specific finding associated with cognitive and functional decline, morbidity and death in a number of neurodegenerative conditions, which raises the question as to what exactly it is that these IIV metrics are measuring. In the current study, we sought to address this question by examining the construct validity and clinical correlates of cognitive IIV in the Texas Alzheimer's Research and Care Consortium (TARCC) Hispanic Longitudinal Cohort (HLC). Methods: The sample included baseline data from 1790 older adults, including 1002 older adult controls (OAC), 287 individuals with subjective cognitive decline (SCD), and 501 individuals with a diagnosis of mild cognitive impairment (MCI). IIV was calculated as the standard deviation of tests across all cognitive tasks (IIV-Across) and between the mean scores of tests within each of four cognitive domains (IIV-Between: attention, language, memory, and executive domains). These IIV metrics were compared between groups and examined in relation to clinical characteristics, biological risk factors, and blood-based biomarkers representing microvascular dysfunction (VEGFR-1: vascular endothelial growth factor receptor-1), neuroinflammation (C-reactive protein), and thrombin generation (FVII: coagulation factor VII). Results: The OAC group demonstrated significantly lower IIV-Across than SCD (p<0.01) and MCI (p<0.001) groups but there were no significant between-group differences for IIV-Between. In the combined sample, greater IIV was found in Hispanic vs. non-Hispanic individuals and in those with anxiety, apathy, and dysphoria, but these effects were stronger for IIV-Across than IIV-Between. Both IIV metrics were related to lower cognitive status and higher CRP values, but only IIV-Across was associated with higher pulse pressure and IIV-Between with higher FVII levels (p<0.001). The two IIV metrics shared a relatively modest correlation with each other (R2=0.56) and with mean performance level in different cognitive domains (R2=0.01-0.03), thereby suggesting that they may be measuring somewhat distinct neuropsychological processes. Conclusion: Findings suggest that different pathophysiological processes hypothesized to occur in the prodromal phase of AD may be reflected in different IIV metrics which appear to have their own strengths and weaknesses, depending on their intended use.

Funding Disclosure: TARCC

Poster #38

COMPARISON OF GENDER DIFFERENCES ON THE QUICK MILD COGNITIVE IMPAIRMENT (QMCI) SCREEN IN COGNITIVELY NORMAL AND MILD COGNITIVE IMPAIRMENT GROUPS

Presenting author: Robin C Hilsabeck, PhD (Associate Professor, The University of Texas at Austin Dell Medical School)

Background: The quick mild cognitive impairment (QMCI) is a relatively new cognitive screening measure that prior research suggests is equal to or better than other screening measures at differentiating individuals who are cognitively normal (CN) from those with mild cognitive impairment (MCI). However, limited research to date has explored the role of demographic factors such as gender on interpreting performance. The aim of our current study was to examine gender differences on the QMCI total and subtest scores in diagnostic groups. It was hypothesized that CN women would perform better than CN men on verbal memory and fluency tasks and that CN men would perform better on the clock drawing task. In contrast, it was also hypothesized that women with MCI would perform worse than men with MCI on verbal memory measures because prior research has shown this performance advantage is no longer apparent once a cognitive disorder is present. Methods: The sample included data from 123 (75 CN, 48 MCI) older adults, 56 men and 67 women. One-way and multivariate analyses of variance using covariates as appropriate were used to investigate group differences in the QMCI total score and six subtest scores. Results: No significant demographic differences between CN and MCI groups were found, but women had significantly less education than men. As expected, CN participants obtained significantly higher QMCI total scores than MCI participants across total and subscales except Word Registration and Clock Drawing (p < .001). There were no statistically significant gender differences on QMCI total or any subtest score in either diagnostic group. While CN women obtained higher scores on Verbal Fluency, Delayed Recall, and Logical Memory and CN men obtained higher scores on Clock Drawing, none of these differences were statistically significant. In the MCI group, women obtained higher Delayed Recall and Logical Memory scores but not Verbal Fluency scores; none of the differences were statistically significant.Conclusion: Findings suggest that gender differences in CN and MCI groups are not statistically significant on the QMCI. However, replication of these findings in larger and more diverse samples are needed before firm conclusions can be drawn.

Funding Disclosure: NIH/NIA R61AG069780

Poster #39

ANALYZING DIRECT EFFECTS OF VARIOUS DEMOGRAPHIC FACTORS BETWEEN HISPANICS AND NON-HISPANIC WHITES ON BOSTON NAMING TEST PERFORMANCE

Presenting author: Paulina Devora, BA (Graduate Student, The University of Texas at Dallas)

The assessment of picture naming is commonly used in both clinical and research settings for diagnosing neurodegenerative disorders such as Mild Cognitive Impairment (MCI) and Alzheimer's Disease (AD). The most widely used measure of confrontation naming is the Boston Naming Test (BNT), but performance can be affected by several demographic factors including age, level and quality of education, gender, and ethnicity. Prior studies have established that healthy Non-Hispanic Whites perform better on the BNT when compared to members of minority groups. However, factors underlying score discrepancies between different ethnic groups have not been fully explored. The current study examines data from cognitively normal Hispanic and Non-Hispanic White adults from the Longitudinal Continuation of Texas Alzheimer's Research and Care Consortium (TARCC) Hispanic Cohort project. Hierarchical regressions to test for ethnic differences between education and estimated VIQ as predictors of BNT were conducted. The modeling involved hierarchically entering sex, age, ethnicity, VIQ, and education, sequentially followed by Ethnicity X Education and Ethnicity X VIQ interaction terms. Ethnicity was observed to be differentially associated with BNT for the two ethnic groups. BNT was more strongly associated with education for Hispanics (R 2= 0.04) compared to Non-Hispanic Whites (R2 = 0.01). Furthermore, BNT also was more strongly associated with VIQ for Hispanics (R2 = 0.15) than for Non-Hispanic Whites (R2 = 0.12). Thus, it can be concluded that education and estimated VIQ are both more strongly associated with BNT performance in Hispanics than in Non-Hispanic Whites, highlighting the importance of properly controlling for these factors when assessing for impairments in cognition.

Funding Disclosure: TARCC

Poster #40

NEUROPSYCHOLOGICAL PROFILE OF AGE-RELATED CHANGES IN MILD INTELLECTUAL DISABILITY: CASE REPORT

Presenting author: Lubnaa Abdullah, PsyD, ABPP (Assistant Professor, University of Texas Rio Grande Valley)

Background: There is available research that highlights the neuropsychological pattern of aging in severe and highly debilitating forms of intellectual disability (ID), such as Down's Syndrome. However, less is known about the neuropsychology of aging and memory in mild or acquired forms of ID. METHOD: The following case illustrates a neurocognitive profile in a 67 year-old, cis-gender, Caucasian male with a previous diagnosis of Major Depressive Disorder (MDD). He complains of lonliness, chronic suicidal ideation and plan with one previous suicide attempt, recent memory changes, confusion, and exploitation by others requiring intervention by Adult Protective Services (APS). RESULT: Results of his neuropsychological evaluation demonstrate an RIAS-2 IQ of 64, and variable performance in verbal, memory and executive functioning. He is diagnosed with mild ID and age-related memory changes. Assessment reveals a complex psychiatric profile associated with severe psychosocial trauma. He is treated with sertraline injections and recommended for social services and milieu therapy. This case will discuss the implications for assessment of abnormal aging in the ID population and the psychiatric functioning of individuals with ID in later life. Recommendations regarding diagnostic considerations will be discussed. CONCLUSION: The assessment of aging in geriatrics with mild forms of ID requires the consideration of many factors that remain sensitive to pre-existing deficits and functioning.

Funding Disclosure: None

Poster #41

EXECUTIVE FUNCTIONING IS AN IMPORTANT PREDICTOR OF LIFE EXPECTANCY IN THOSE WITH ALL-CAUSE DEMENTIA

Presenting author: Jeff Schaffert, PhD (Assistant Professor, UT Southwestern Medical Center)

Background: Neuropsychiatric, functional, motor, and demographic factors have been associated with life expectancy (LE) in those with dementia. Recent findings suggested cognition (assessed by the Mini-Mental State Exam [MMSE]) to be the strongest predictor of LE in Alzheimer's disease (AD; Schaffert et al., 2022). We evaluated if more detailed neuropsychological scores predict LE in a larger sample of individuals with all-cause dementia. Method: Participants were 4,090 deceased individuals (Mean age=74.5, Mean education=14.8, Male=44%, White=90%, Non-Hispanic=96%) with all-cause dementia (at visit 1, AD=78%) from the National Alzheimer's Coordinating Center. Three index scores [executive function/speed/attention (EFAS), language, memory] were calculated from the NACC neuropsychological batteries. Variables (from visit 1) were entered into a forward regression model (p<.001 as point-of-entry) to predict days of LE, and included: age, gender, race (white/non-white), ethnicity (Hispanic/Non-Hispanic), diagnosis (AD/non-AD), abnormal neurological exam (yes/no), Functional Activities Questionnaire (FAQ, total score), Neuropsychiatric Inventory Questionnaire (NPI-Q, total score), MMSE, and EFAS, language, and memory composite Z-scores. Results: Performance on the EFAS composite explained the most variance in LE (R2=.065), followed by age (R2=.044), diagnosis (R2=.023), FAQ (R2=.016), gender (R2=.012), abnormal neurological exam (R2=.006), NPI-Q (R2=.004), language abilities (R2=.003), and Hispanic ethnicity (R2=.003). Plus/minus one Z-score on the EFAS composite predicted 158 days of LE, and each year of age predicted 27 days of LE. Conclusions: EFAS performance and age explained >10% of LE variance. The MMSE failed to predict LE in this model that included more detailed neuropsychological data. EFAS impairment may be a more important predictor of LE compared to other neurocognitive domains and cognitive screeners.

Funding Disclosure: None

Poster #42

THE ASSOCIATION OF NEIGHBORHOOD TRUST AND VIOLENCE WITH WHITE MATTER INTEGRITY IN A DIVERSE SAMPLE

Presenting author: Anthony Joseph Longoria, MS (Graduate Student, UT Southwestern Medical Center)

Objective: Hispanic and non-Hispanic Black older adults are at increased risk of dementia compared with non-Hispanic Whites. Cerebral white matter disease may contribute to this disparity, though little is known about how factors such as neighborhood disadvantage and perceived economic status may relate to white matter integrity in diverse groups. Method: Participants from a population-based, multiethnic cohort (N=2766; Female=58%; Black=48%; Hispanic=14%) completed a measure of neighborhood perceptions ("resources," "trust," "violence") and perceived SES ("access to food," "access to healthcare") and underwent 3T brain MRI with 2D FLAIR and 3D T1-weighted sequences to assess white matter hyperintensity (WMH) burden and white matter volume (WMV), respectively. Linear regression models examined associations among these variables controlling for demographic and traditional SES variables (income, education), followed by post-hoc analyses by race/ethnicity and sex. Results: Reported income, education, ethnoracial group, and sex were associated with WMH in the overall sample (ps<.05), and "trust," "access to healthcare," income, education, race/ethnicity, and sex were significantly associated with cerebral WMV. Post-hoc analyses revealed "violence" was associated with more WMH in Black women (p<.05,β=-0.12), lower "trust" was associated with lower WMV in Hispanic men (p<.05,β=0.22), and lower "access to medical care" was associated with lower WMV in White women (p<.05,β=-0.18). Conclusions: Neighborhood disadvantage and perceived SES were significantly associated with indicators of white matter pathology, with differential relationships across race and sex. Findings may reflect disproportionate environmental and social stressors encountered by some groups. Future analyses will examine whether these sociocultural factors relate to white matter burden and cognitive function later in life.

Funding Disclosure: Moss Heart Trust

Poster #43

ASSESSING THE EXTERNAL VALIDITY OF A MAGNETOENCEPHALOGRAPHY-BASED VERBAL WORKING MEMORY TASK IN COGNITIVELY HEALTHY ADULTS

Presenting author: Sahil Chilukuri, BS (Research Intern, UT Southwestern Medical Center)

Introduction: Verbal working memory (VWM) encompasses the temporary maintenance and/or manipulation of verbal information to be used towards concurrent processing, and in prior studies we have developed and validated a Sternberg- type VWM task to be performed during magnetoencephalographic (MEG) neuroimaging. However, how performance on the VWM task relates to performance on other well-established neuropsychological instruments, that are commonly administered in both clinical and research settings, is unknown. The goal of the present investigation was to assess the external validity of the MEG-based VWM task. Methods: Fifty cognitively healthy adults (29 Female, M age = 59.04) performed the VWM task during MEG. During the task, participants were presented with sets of 6 letters (encoding) and, following a brief delay (maintenance), responded as to whether a test letter was in the previous 6 letters (retrieval). 128 trials were collected and average reaction time across correct trials (ms) and accuracy (% correct) were computed per participant. Participants also completed an animal fluency task (AF), the Weschler test for adult reading (WTAR), and the Montreal Cognitive Assessment (MoCA) outside of the scanner. Partial correlations were computed to assess the relationships between VWM performance (accuracy, reaction time) and neuropsychological performance (AF, WTAR, MoCA), while controlling for age, and a Bonferroni correction was applied. Results: There was a significant relationship found between VWM accuracy and performance on the WTAR (pr = 0.487, p<0.001) and MoCA (pr= 0.455, p= 0.001). VWM reaction time was not significantly related to neuropsychological task performance. Conclusion: The external validity of the VWM task (the extent to which task results are generalizable to real world situations) has been demonstrated to a certain degree, but further analysis which incorporates a broader range of neuropsychological metrics is necessary. In future analyses, the MEG-derived neural oscillatory dynamics underlying VWM-task performance should also be utilized to investigate neurobehavioral relationships.

Funding Disclosure: None

Poster #44

PRELIMINARY EXPLORATION OF A NOVEL SPEECH ANALYSIS ALGORITHM TO DETECT COGNITIVE IMPAIRMENT IN A SPANISH POPULATION

Presenting author: Alyssa Kaser, BA (Graduate Student, UT Southwestern Medical Center)

Background: Early detection of mild cognitive impairment (MCI) and dementia is crucial for initiation of treatment and access to appropriate care. While comprehensive neuropsychological assessment is often an intrinsic part of the diagnostic process, access to services may be limited and cannot be utilized effectively on a large scale. For these reasons, novel technologies to screen for cognitive changes in older individuals are evolving as new avenues for early detection. The present study presents preliminary data on a new technology that uses automated linguistic analysis software to screen for MCI and dementia. Methods: Data were collected from 148 Spanish-speaking individuals recruited in Spain (MAge=74.4, MEducation=12.93, 56.7% females) of whom 78 were diagnosed as cognitively normal [CN; MMMSE = 28.51 (1.39)], 49 as MCI [MMMSE = 25.65 (2.94)], and 21 as all-cause dementia [MMMSE = 22.52 (2.06)]. Participants were recorded performing various verbal tasks [Animal fluency, phonemic (F) fluency, Cookie Theft Description, and CERAD list learning task]. Recordings were processed via text-transcription and sound signal processing techniques. Receiver Operating Characteristic (ROC) analysis was conducted to determine sensitivity and specificity of predicted algorithm performance [CN vs. impaired (MCI or dementia)] against clinical diagnoses. General linear modeling was used to test whether age, sex, education, and multilingualism significantly predicted logistically transformed weighted algorithm scores. Results: Scores were transformed to logit scores, with significant differences in mean logit scores between all groups (p<.001). Logit-inverse transformation of mean logit scores (possible range 0 -1) resulted in values of 0.06 for CN, 0.90 for MCI, and 0.99 for all-cause dementia groups. ROC curve analyses revealed the algorithm obtained a total area under the curve of 0.92, with an overall accuracy of 86.8%, a sensitivity of 0.92, and specificity of 0.82. Age was identified as a significant predictor (beta = 0.22; p <0.01) of algorithm output. Conclusions: Findings provide initial support for the utility of an automated speech analysis algorithm to detect cognitive impairment quickly and efficiently in a Spanish-speaking population. Additional research is needed to validate this novel methodology in other languages, as this may represent a promising cross-cultural screening method for MCI and dementia detection.

Funding Disclosure: AcceXible Impacto, Sociedad Limitada

Poster #45

NOT NORMAL BUT NOT MCI: COURSE OF MEMORY OVER TIME

Presenting author: Michael Conley, MCRC (Graduate Student, UT Southwestern Medical Center)

Objective: A diagnosis of mild cognitive impairment (MCI) requires memory complaint and objective memory impairment. This study aimed to compare memory performances over time in individuals who do not meet traditional MCI criteria to those with normal cognition (NC) and those who converted to MCI. Participants and Methods: 827 participants from a longitudinal sample of the Texas Alzheimer's Research and Care Consortium were classified by retrospective diagnosis: 1) NC over 5 visits (n=511, 71% female; 42% Latinx/Hispanic), 2) baseline NC to amnestic MCI (n=62; 63% female; 57% Latinx/Hispanic), 3) subjective memory complaint (SMC) at last visit (n=133; 58% female; 70% Latinx/Hispanic), and 4) impaired but not MCI (i.e., no complaint) at last visit (n=121; 71% female; 60% Latinx/Hispanic). A memory composite (z-score) was created from the CERAD list-learning task (immediate, delayed, and recognition-discrimination) and Wechsler Memory Scale (Immediate and Delayed Logical Memory and Visual Reproduction) to evaluate memory performance over time. Linear mixed-models adjusting for age, education, sex, ethnicity, and number of APOE e4 alleles evaluated memory performance and depression over time. Results: At baseline, groups differed by age (F=22.82; p<.001), education (F=8.60; p<.001), MMSE scores (F=9.38; p<.001), depression (GDS-30) (F=3.56; p=.015), and memory performance (F=24.29; p<.001). A significant group X time interaction was observed (F=4.83, p<.001). Memory improved in both the SMC and NC groups, remained stable in the impaired but not MCI group, and declined in those who converted to amnestic MCI. Depression scores also showed a significant group X time interaction (F=2.43; p=.004), in which the NC to MCI group endorsed slightly more depression symptoms over time, while other groups declined or remained stable. Conclusions: Memory trajectories in this diverse sample differed across groups. Individuals with SMC but without objective memory impairment and normal controls showed some improvement in memory over time, presumably due to practice effects. Those with subtle memory impairments but no complaint (i.e., did not meet MCI criteria) remained stable and those who converted to amnestic MCI had worse memory across time. The stability of memory performances in the impaired not MCI group suggests these subtle memory inefficiencies may be longstanding or unperceived.

Funding Disclosure: TARCC

Poster #46

ASSESSING CHALLENGES OF CAREGIVERS OF ELDERLY PEOPLE WITH DEMENTIA IN TEXAS TO ESTABLISH A DIGITAL PLATFORM: A QUALITATIVE STUDY

Presenting author: Jeswin Vennatt, MBA (Medical Student, Texas A&M University Health Science Center)

Background: The caregivers of people living with Alzheimer's disease and dementia face a high degree of psychological, physical, and financial stress. However, the needs of caregivers of people with dementia are poorly understood. The objectives of the study are to identify the living, financial, and legal challenges faced by caregivers of people living with dementia, and to understand their expected features and needs in order to develop a digital platform. Method: A qualitative study was conducted among unpaid caregivers for people with dementia in Texas. Caregivers who were 1) non-paid caregivers for dementia patients, 2) actively involved in the care recipient's living, legal or financial decisions, and 3) adults (18 years and older) were included in the study. A one-on-one semi-structured interview was conducted via Zoom with each participant. The interviews are transcribed and analyzed using NVivo (Version 12.0) for thematic analysis. Results: The pre-screening survey was completed by 323 respondents, and 28 of them met the eligibility criteria for the interview. According to survey responses, 64.6% of respondents reported challenges in making financial or legal decisions for their loved ones. The main challenges reported by caregivers included balancing other family obligations and work, financial burdens, finding reliable providers and services, providing transportation, tension with family members, navigating the legal process, coping with catastrophic events, and lack of mental health support. The digital platform is expected to include support groups, education materials, communication with doctors, psychologists, psychiatrists, or licensed professional counselors to assist caregivers, peer-reviewed research evidence about dementia, task-based reminders (such as taking medication), and a comprehensive database to locate reliable services. Conclusion: Caregivers of people with dementia face a variety of living, financial, and legal challenges. Innovative digital platforms and community participatory initiatives need to be developed in order to provide caregivers with support. These innovations can empower caregivers and result in the successful aging of their loved ones.

Funding Disclosure: NIA Small Business Innovation Research Grant (SBIR) Grant

Poster #47

CAREGIVER-REPORTED DECLINES IN ACTIVITIES OF DAILY LIVING AMONG PERSONS WITH MILD DEMENTIA DURING THE COVID-19 PANDEMIC

Presenting author: Matthew Lee Smith, PhD, MPH, CHES (Professor, Texas A&M University)

Background: Although cognitive and physical functioning are expected to decline over time among persons living with dementia, the COVID-19 pandemic may have accelerated these declines because of required shut-downs, stay-at-home orders, and service disruptions. This study examines the naturally occurring declines in activities of daily living (ADL) among persons with mild dementia during the early stages of the COVID-19 pandemic. Methods. As part of an NIH-funded study, caregivers (CG) were asked to manually assess their care recipients (CR) each month using the Alzheimer's Disease Cooperative Study-Activities of Daily Living Scale (ADCS-ADL). Assessments took place from October 2020 to August 2022, and the maximum study duration was 18 months. The ADCS-ADL consists of 23 items, with 6 basic ADL (BADL) items and 17 instrumental ADL (IADL) items. Analyses included 110 CR. Cox regression models were fitted to time to 25% score decreases in the total ADCS-ADL, BADL, and IADL over the study period. Each regression model controlled for the CR's age, sex, and baseline Mini-Mental State Examination (MMSE) score. Sensitivity analyses were performed based on if the CR lived with their CG. Results: The majority of CR was female (57.3%) and lived with their CG (56.4%). On average, CR were age 78.7 (±8.5) years, had a baseline MMSE of 22.6 (±2.7), and had a baseline ADCS-ADL of 56.0 (±16.3). On average, a one unit decrease in baseline ADCS-ADL corresponded to a 3% hazard rate increase for a 25% decline in ADCS-ADL over time (HR=0.973, P=0.006). For BADL and IADL, a one unit decrease in baseline MMSE corresponded to a 14% (HR=0.876, P=0.021) and 13% (HR=0.883, P=0.027) hazard rate increase for a 25% decline over time, respectively. In sensitivity analyses, CR living with their CG reported more dramatic declines in total ADCS-ADL, BADL, and IADL scores compared to CR not living with their CG, respectively. Conclusions: This study investigates factors associated with rates of ADL decline among persons with mild dementia during the COVID-19 pandemic. Findings suggest that CG residing with CR and baseline MMSE are useful for the timely initiation of additional care-related supports and services.

Funding Disclosure: This study was funded by the National Institute on Aging (NIA) as part of an SBIR (#R44AG065118-03).

Poster #48

UTILIZATION AND PERCEIVED USEFULNESS OF MONITORING TECHNOLOGY FOR FAMILY CAREGIVERS OF PEOPLE LIVING WITH DEMENTIA

Presenting author: Matthew Lee Smith, PhD, MPH, CHES (Professor, Texas A&M University)

Despite the proliferation of technology-based solutions for caregiving, less is known about their utilization and perceived value among caregivers of persons living with dementia (PLWD). In collaboration with three TARCC sites (Texas A&M, University of North Texas, and University of Texas at Austin), PLWD and family caregiver dyads were recruited from clinical and community sites. PLWDs were asked to wear a GPS-based wearable device, which were paired with caregivers' smartphone application that enabled location monitoring. The wearable device was equipped with an accelerometer and call functions. After three months, researchers called caregivers to ask about their utilization of the system (i.e., wearable device paired with smartphone application) and the perceived value of this caregiving technology. Forty-one caregivers completed follow-up telephone interviews. About 70% of caregivers reported their care recipient wore the wearable device daily, and 39.1% of caregivers used the smartphone application daily. Approximately 31% of caregivers reported daily use of the tracking feature, 30.8% reported daily use of the "safe zone" feature (i.e., geo-fencing), and 17.1% reported daily use of the two-way calling feature. About 39% of caregivers were extremely satisfied with the system, 43.6% found the system extremely easy to use, and 46.2% found the system extremely useful for caregiving. On average, caregivers with higher baseline Zarit Burden Interview scores were more satisfied with the system (f=3.75, P=0.034) and found the system to be more useful with their caregiving (f=5.97, P=0.006). Findings suggest that caregiver burden may drive the perceived usefulness of, and satisfaction with, technology-based solutions.

Funding Disclosure: TARCC

Poster #49

GENDER DIFFERENCES IN CARE NETWORKS OF OLDER ADULTS WITH DEMENTIA

Presenting author: Phillip A. Cantu, PhD (Assistant Professor, University of Texas Medical Branch at Galveston)

Background: Gender differences in care networks, the use of paid and informal caregivers, and the need for care among older adults with dementia remain unclear. Objective: To evaluate gender differences in care networks of community dwelling older adults with dementia in the United States. Methods: We used data from the 2018 wave of the Health and Retirement Survey, a nationally representative sample of older adults in the United States. We included individuals aged 70 and older with dementia. We estimated the total annual care hours received from paid and unpaid informal caregivers. Paid care was categorized as formal institutional care, paid care from family and friends, and care paid by insurance. Informal care was categorized as from a spouse, children, or other sources. We used OLS regression to predict gender differences in types of care controlling for demographic characteristics, wealth, and level of disability. Results: Our sample was 61.3% women, with a mean age of 85.1. Women received on average 1,384 total hours of care a year compared to 1,015 hours for men, with a greater proportion of women's care coming from paid sources. Differences in hours of care from all sources were explained by higher levels of disability for women, except for differences in hours of care from spouses and children. Conclusions: There were more women than men with dementia and, as a result, they had higher levels of disability. The greater cost of care for older women is reflected by more hours of formal and informal care.

Funding Disclosure: K12HD052023 P30AG059300 P30AG059301

Poster #50

FEASIBILITY OF THE CART IN-HOME TECHNOLOGY PLATFORM IN A PILOT STUDY OF HISPANIC OLDER ADULTS RESIDING IN SOUTH TEXAS

Presenting author: Julia Mathews, BS (Graduate Student, UT Health Science Center San Antonio)

Background: As the prevalence of Alzheimer's disease (AD) rises, the development of effective screening tools are needed to aid in earlier AD diagnosis. Early identification is important to help maintain patient and caregiver quality of life, as well as to improve patient care. In the USA, Hispanic and Non-Hispanic Black adults are at a disproportionately higher risk of developing dementia than non-Hispanic whites, but are poorly represented in AD research. Researchers at OHSU developed the Collaborative Aging Research using Technology (CART) platform, which has the potential to remotely and continuously detect early declines in cognition by analyzing changes in everyday activity levels. Although the platform is promising, installing digital technology in a home can cause concerns about privacy and user friendliness, especially in older adults. In our CART pilot study at UTHSCSA, we sought to determine the feasibility of the CART platform in an underrepresented, mostly Hispanic population and to gain insight from their experiences. Method: A total of seven participants were enrolled in our year-long pilot study. The CART technology platform consisted of a wearable watch, digital scale, electronic pillbox, bed mat (to measure sleep), and wall-mounted movement sensors, as well as a small computer to upload device data to secure servers. Participants completed the UDS3 neuropsychological battery at baseline and the one-year end of study. Participants also completed an end of study user experiences survey. Result: Our cohort consisted of five Hispanic, one non-Hispanic Black, and one non-Hispanic white participant, ages 65-80. There were no significant changes in cognition over the course of the study (all p>0.05); however, self-reported depressive symptoms reduced (p=0.034). The responses to the user experience survey demonstrated that our participants found the technology installation satisfactory (43%) or very satisfactory (57%) and perceived that their privacy concerns were addressed satisfactorily (43%) or very satisfactorily (57%). Conclusion: Our results suggest that the CART platform was widely accepted by our mostly Hispanic older adult participants and that they were comfortable with the study devices. The findings support the feasibility of increasing representation of diverse older adults in dementia prevention research using digital devices.

Funding Disclosure: NIH P30AG066546

Poster #51

VALIDATION OF THE MULTIDIMENSIONAL HEALTH PERCEPTIONS QUESTIONNAIRE IN CARE PARTNERS OF INDIVIDUALS WITH ALZHEIMER'S DISEASE AND RELATED DEMENTIAS (ADRD)

Presenting author: Alexandra B Holland, LMSW (Psychological Associate, UT Southwestern Medical Center)

Background: Health beliefs affect healthcare engagement, treatment adherence and uptake, and clinical outcomes. Understanding these beliefs is a critical component of person- and family-centered care. We developed the Multidimensional Health Perceptions Questionnaire (MHPQ) as an all-encompassing measure of health perceptions. The MHPQ has seven health perceptions subscales validated among both English and Spanish-speakers in the general population (content validity index of 98.1%; Cronbach's alphas >.70 for all subscales). Our aim is to describe the health perceptions of care partners of individuals with Alzheimer's Disease and Related Dementias (ADRD) participating in our Care Partners in Dementia Problem Solving (CaDeS) trial, a TARCC collaborative project. Method: Care partners of individuals with ADRD participating in CaDeS completed the MHPQ. It includes 65 items rated on a 1-to-5 agreement scale with items and instructions at a modified SMOG reading level of 6th grade in English and 8th grade in Spanish. It has 7 subscales yielding average scores (1-5): (1) anticipated discrimination and judgment, (2) spiritual health beliefs, (3) social and emotional wellbeing beliefs, (4) confidence and trust in healthcare providers and medicine, (5) health self-efficacy/internal locus of control, (6) trust in social health advice, and (7) health literacy. Results: N=71 care partners are included in this analysis. Internal consistency reliabilities were acceptable to good for all but 2 subscales (Cronbach's alphas= .748-.856; .584-.682). Means and standard deviations for all subscales were as follows: (1) 1.9155 (SD .5259); (2) 2.685 (SD .63093); (3) 3.518 (SD .41274); (4) 3.5813 (SD .37023); (5) 3.9525 (SD .44036); (6) 2.6268 (SD .53264); (7) 4.2394 (SD .43254). We will further present differences in health perceptions by race, ethnicity, language, age, and relationship (e.g., spouse, child, etc.) Conclusion: By understanding care partner health perceptions and how this affects their health outcomes, healthcare providers can adapt their intervention approaches for an increasingly diverse population and implement patient/family-centered care. We plan to examine how differences in health perceptions affect engagement, uptake, and outcomes after Problem-Solving Training (PST), at the completion of the CaDeS trial. These data will inform important adaptations to PST that may be required based on care partners' health perceptions.

Funding Disclosure: TARCC

Poster #52

MACHINE LEARNING ALGORITHM TO PREDICT DURATION TO FULL TIME CARE AFTER ALZHEIMER'S DISEASE DIAGNOSIS

Presenting author: Jessica Helphrey (Graduate Student, UT Southwestern Medical Center)

Background: Patients and their families often ask clinicians to estimate when full-time care (FTC) will be needed after Alzheimer's Disease (AD) is diagnosed. Our objective was to develop a clinically relevant, data-driven predictive model, that provides precision and user-friendliness, using machine learning to estimate time to FTC in AD based on information gathered from clinical interview plus neuropsychological data. Method: The National Alzheimer's Coordinating Center dataset was used to examine 3,809 participants (M age at AD diagnosis = 76.05, SD = 9.76; 47.10% male; 87.20% Caucasian) with AD dementia aged ≥50 years, had no history of stroke, and not dependent on others for basic activities of daily living at time of diagnosis. To develop a predictive model for time until FTC, supervised machine learning algorithms were implemented. In Model 1, 29 variables captured at the time of AD diagnosis were included. In Model 2, additional neuropsychological variables were added. Algorithms were trained and tested, cross validation was conducted using bootstrapping, and accuracy and precision calculations were performed. Result: The average time to requiring FTC after AD diagnosis was 3.32 years. For the clinical interview model (Model 1), younger age of onset, use of cholinesterase inhibitor medication, incontinence, apathy, living alone, a positive family history of dementia, and lower MMSE score significantly predicted duration to FTC. In the clinical interview + neuropsychological variables model (Model 2), the clinical predictors remained significant, and lower Boston Naming Test and Digit-Symbol Coding scores showed the largest effects in predicting duration to FTC. After testing prediction models, the average estimated time to FTC using the clinical interview model was within an average of 5.2 months of the recorded event and within an average of 5.8 months for the model with neuropsychological data. Conclusion: Predicting when individuals diagnosed with AD will need FTC is important as the transition often carries significant financial costs related to caregiving. Duration to FTC was predicted by clinical and neuropsychological variables that are easily obtained during standard dementia evaluations. Implementation of the model for prediction of FTC in cases showed encouraging prognostic accuracy.

Funding Disclosure: AWARE foundation

Poster #53

A PUBLIC HEALTH CRISIS IN RURAL WEST TEXAS: NEUROCOGNITIVE DISPARITIES BETWEEN HISPANIC AND NON-HISPANIC COMMUNITIES MODERATED BY DEPRESSION

Presenting author: Carol Fadalla, BA (Graduate Student, Texas Tech University)

Background: Hispanic adults are at increased risk for neurocognitive decline compared to other racial-ethnic minorities. Few studies have examined the prevalence of neurocognitive decline among Hispanics living in rural West Texas-a population deeply underserved. The study explored whether depression moderated prevalence utilizing data from Project FRONTIER (Facing Rural Obstacles to Healthcare Now Through Intervention, Education, & Research). Methods: 1,842 participants (Mage = 59.68 years, SDage = 12.21) 1,053 Hispanics and 752 non-Hispanics, from multiple rural counties in West Texas between June 2006 and January 2022. We evaluated neuropsychological testing using five measures: The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), The Trails Making Test A and B, and Clock Drawing 1 and 2. Geriatric Depression Scale was used to assess depression and we dichotomized using a validated cutoff of "no depression" (0-10) and "mild/severe depression" (11-30). Results: The MANOVA identified a significant interaction between ethnicity and depression on neurocognitive functioning (Roy's Largest Root = 3.56, F(5, 1,548), p= 0.003, η2p = 0.01). More specifically, non-Hispanic individuals who reported mild/severe depression had worse neurocognitive functioning compared to participants with no depression. However, there was no difference for Hispanic participants who reported mild/severe depression compared to no depression. A post-hoc ANOVA revealed a significant main effect on neurocognitive functioning (RBANDS) for ethnicity (F(5, 1,549), p < .001, η2p = 0.06), as participants who identified as Hispanic (M= 79.41 and SD= 0.57) had lower overall neurocognitive functioning than non-Hispanic (M= 88.38; SD= 0.78) participants. On indicators of neurocognitive functioning, a significant main effect for depression was found (F(5, 1,549) = 13.62, p < 0.001, η2p = 0.04). Conclusion: Results are consistent with the minority stress model, which states racial-ethnic minorities are exposed to chronic stressors, which negatively influences functioning. Non-Hispanic individuals encounter fewer chronic stressors than minorities and, the introduction of stressors, like depression, may impact neurocognitive functioning-unlike Hispanic individuals. Previous studies suggest chronic life stressors are equally detrimental to cognitive functioning in individuals with and without depression. Results support the need for culturally tailored interventions because underlying factors related to cognitive function differ, such as depression in non-minority individuals.

Funding Disclosure: None

Poster #54

NEIGHBORHOOD PHYSICAL DISORDER AND ARCHITECTURAL FEATURES OF STREETSCAPES OF PARTICIPANTS OF THE LONGITUDINAL HISPANIC COHORT IN CAMERON COUNTY: PRELIMINARY RESULTS

Presenting author: Cristian Maestre-Nunez, BArch (Graduate Student, University of Texas Rio Grande Valley)

Background: An estimated 23% of global deaths are due to the environments in which people live. We aim to characterize the level and characteristic of neighborhood physical disorder in areas where participants of the Longitudinal Hispanic Cohort reside in the Rio Grande Valley, Texas. In addition we will characterize prevalent architectural features in streetscapes of one of the counties, Cameron, Texas. Methods: We used Google Street View Imagery to rate elements of the neighborhood physical disorder score in the Rio Grande Valley. In addition we studied different architectural attributes of the streetscape in Cameron County. Elements analyzed included: site connectivity with context, pedestrian quality, scale of buildings, architectural elements, landscaping, and signage. Results: We generated a map to visualize the distribution and intensity of neighborhood physical disorder in the Rio Grande Valley. Results indicate that the distribution is heterogeneous within each of the cities and between cities. Some of the most frequent architectural features in streetscapes are: -There is little interrelationship between open spaces and buildings. Parking lots are usually in front of buildings. -Urban streetscapes are mostly designed for motorists and not for pedestrians. -Most buildings are at human scale, with few buildings over 5 floors. -The most common residential architectural style is Ranch-Style house characterized by single-story, open space concept, almost non-existent ornaments, and low-pitch rooflines. -The tall palms characteristic of the region add scale to street edges; however, their canopies are high and small, and few provide shade. -Blank walls are common, with few exhibiting architectural features that enhance streetscapes. -Most signage is commercial and does not enhance building character or the pedestrian experience. Discussion: Neighborhood physical disorder is high and heterogenous in the area. We identified the most frequent typologies of architectural design in Cameron County. This information will be particularly relevant for community advocates and city planners as well as to residents and researchers interested in understanding influences on urban health.

Funding Disclosure: NIH DP1AG069870, P30AG059305 and P30AG066546

Poster #55

SERVICE USAGE AMONG CARE PARTNERS OF ADULTS WITH DEMENTIA IN THE CADES TRIAL

Presenting author: Jill Morales, MS (Research Study Coordinator, UT Southwestern Medical Center)

Background: Participants in the Care Partners in Dementia Problem-Training (CaDeS) study are care partners who are involved in the care of a loved one with Alzheimer's Disease or related dementias (ADRD). Care partners include parents, adult children, other family members, or close friends. They perform many jobs to assist their loved ones; however, they may need the assistance of community resources (e.g., home care, adult daycare, etc.). Here, we examine whether the care partner's relationship type (e.g., spouse, adult child, etc.) was associated with the types of community resources and services they used. Method: Care partners completed an initial evaluation that included information about their demographics and other personal information. They were asked to report their relationship with the person with ADRD and rate the quality of that relationship on a 10-point scale; higher ratings indicate higher relationship quality. They were also asked about community resources and services used as a care partner. Result: Eighty-two care partners were included in analyses. Care partner relationships with the care recipient were as follows: n=44 spouses, n=19 parents, n=13 adult children, n=4 other family member, and n=2 siblings. Over 90% of care partners indicated that they used adult daycare, caregiving education programs, written resource materials, respite care, support groups, and/or help from their family or friends. All children (n=13) used these community resources. Spouses, parents, and children reported that they also used community resources such as Meals on Wheels, counseling, care managers, and other community programs available to them. Conclusion: Care partners perform many services for their loved ones, and can benefit from various community resources. While almost all care partners in the CaDeS trial used community resources, adult children care partners seemed to especially rely on these resources. This is consistent with other studies indicating that adult children care partners may experience more feelings of burden and distress than other family caregivers. Unfortunately, not everyone has resources available to them; therefore, future efforts are needed to identify what resources are most beneficial to care partners so that service access can be prioritized.

Funding Disclosure: TARCC

Poster #56

REDUCED PREVALENCE OF DEMENTIA IN PATIENTS TREATED WITH BRAIN-PENETRANT CALCINEURIN INHIBITION

Presenting author: Jacqueline Silva, BS (Graduate Student, University of Texas Medical Branch at Galveston)

Background: Evidence suggests that patients treated with calcineurin inhibitors (CNIs) have a lower prevalence of dementia, including Alzheimer's disease (AD), compared to the general population. While plausible, whether the observed effects are related to brain penetrance remains unresolved. To address this question, we interrogated electronic medical records comparing dementia prevalence in patients treated with the brain penetrant CNI tacrolimus to patients treated with the non-brain penetrant CNI cyclosporine. Methods: We conducted a retrospective cohort study using the TriNetX global health research network. Patients currently over age 65 were included based on prescription of tacrolimus or cyclosporine. Patients who were prescribed both drugs were excluded. Cohorts were propensity-score matched by age, race, sex, and a range of other covariates. The outcomes examined were diagnosis of dementia, including AD, at least thirty days following earliest recorded drug treatment. Results: Of the 65,599 patients in each cohort, 2,587 (4.05%) of those prescribed tacrolimus and 2,929 (4.62%) of those prescribed cyclosporine were later diagnosed with dementia. Tacrolimus treatment was associated with a 12% risk reduction of developing dementia (relative risk of 0.88 of dementia in tacrolimus vs. cyclosporine, p<0.0001). Similar results were achieved when examining AD diagnoses separately (relative risk of 0.71 of AD in tacrolimus vs. cyclosporine, p<0.0001). Conclusion: The results suggest the brain penetrant CNI is associated with a lower prevalence of dementia, including AD, relative to the non-brain penetrant CNI. These data encourage clinical evaluation of localized delivery of tacrolimus to the brain as a viable therapeutic for prevention and treatment of dementia, including AD, with localized delivery potentially reducing systemic side effects.

Funding Disclosure: NIH R01AG060718

Poster #57

GAIT AND MOTOR CHANGES AS PREDICTORS OF ALZHEIMER 'S DISEASE IN INFECTED APP/PS1 MICE

Presenting author: Jocelin Reyes (Undergraduate Student, Texas A&M University)

Background: Aβ plaques and neuronal loss in regions of the brain responsible for cognition, such as the hippocampus and prefrontal cortex, characterize Alzheimer 's Disease (AD). Cognitive symptoms are commonly used to diagnose clinical AD, though recent studies show that changes in motor abilities may occur early in the disease, as AD pathology also develops in the motor cortices, striatum and substantia nigra. Understanding the projection of development of early deficits and determining if motor impairments are part of this early phenotype can aid in prediction of further decline in AD patients who do not yet present detectable cognitive impairment. Further, external insults such as minor chronic infection may be necessary to initiate the disease cascade, and may shed light into the mechanism of disease origination. We hypothesize that a low-grade nonpathological staph infection accelerates the disease process in the APP/PS1 mouse AD model. Methods: Male and female APP/PS1 and non transgenic mice were infected with Staphylococcus aureus (or PBS for control) via nasal inoculation starting at 12 weeks of age twice a week for 3 months. At 6 months of age, mice were tested for motor performance in the rotarod and DigiGait. Results: We observed no differences between vehicle groups of both genotypes. However, Staph infection impaired gait measures in APP/PS1 male mice. The same cohort was also impaired in the rotarod measures. Conclusion: Bacterial infection resulted in lower motor performance in the APP/PS1 mouse model, suggesting that added external insults may be necessary for development of disease-related early deficits.

Funding Disclosure: None

Poster #58

CONTINUOUS WAVE 670NM LED LIGHT THERAPY IMPROVES SHORT- AND LONG-TERM OBJECT RECOGNITION, LOCOMOTION, AND REDUCES PHOSPHORYLATED TAU BURDEN IN 3XTGAD

Presenting author: Kevin Johnson, BS (Graduate Student, University of Texas Medical Branch at Galveston)

Non-invasive therapeutic alternatives for Alzheimer's Disease (AD) are highly sought after due to their accessible nature and sustainability of use. Photobiomodulation (PBM) therapy uses 620-1100nm red to near-infrared light to modulate a variety of biological processes including neuroinflammation, amyloid and tau oligomerization, and apoptosis. 670nm red light has been shown to reduce synaptic vulnerability to amyloid-β oligomers in wild-type (WT) mice and reduce tau oligomerization in multiple mouse models of tauopathy. The 3xTgAD mouse model of AD neuropathology, expressing human APP, PS1, and tau mutations, develops an abundance of amyloid plaques and neurofibrillary tangles causing progressive cognitive dysfunction starting around 12 months of age. Here we sought to determine if 670nm light therapy could rescue cognitive and motor dysfunction and reduce tau hyperphosphorylation in 14-to-18-month-old 3xTgAD mice. METHODS: 14-to-18-month-old 3xTgAD mice were treated with continuous-wave 670nm LED light applied directly to the head for 90s per day, 5 days per week for 4 weeks. 3xTgAD sham and WT control animals were held under the LED device with the light turned off for the same amount of time. During the fourth week of treatment, spontaneous locomotor behavior was measured by placing animals in an empty chamber for 10 minutes and using ANY-maze software to track distance traveled, mobile time, mobile episodes, and latency to mobile episodes. Animals were then tested for working memory in the Y maze spontaneous alternation task and short-and long-term memory in the novel object recognition task during the last 2 days of treatment. At the completion of behavioral testing on treatment day 20, brains were collected for western blot analysis of total tau using the Tau5 antibody (Abcam) and phospho-tau (pTau) using the paired helical filament (PHF13.6) antibody (ThermoFisher). RESULTS: Light-treated 3xTgAD mice showed increased distance traveled compared to sham (17.43 ± 3.46 vs. 9.97 ± 3.03, one-way ANOVA p=0.03), and decreased immobile episodes (4.50 ± 1.92 vs 10.00 ± 1.83, one-way ANOVA p=0.02) in the spontaneous locomotor activity test. Light-treated animals also showed significant object recognition at 2 hours (DI 32.75 ± 7.71, one sample t-test p<0.001) and 24 hours (DI 44.14 ± 15.73, one sample t-test p=0.001) post-training.

Funding Disclosure: NIH R01AG06943302

Poster #59

SELECTIVE UPREGULATION OF THE MASTER REGULATOR E2F1 IN THE CEREBRAL VASCULATURE IMPROVES MEMORY IN AGED MICE

Presenting author: Jose Felix Moruno Manchon, PhD (Instructor, UT Health Science Center Houston)

Background: Brain aging is a major risk factor for the progression of Alzheimer's disease (AD) and Vascular contribution to cognitive impairment and dementia (VCID). VCID is identified by cognitive deficits caused by dysfunction of the cerebral endothelium and disruption of the blood brain barrier (BBB). With aging, cerebral endothelial cells (CEC) enter into a non-proliferative state of senescence characterized by DNA damage and a pro-inflammatory secretory phenotype that negatively affects the cerebral vasculature and central nervous system function. The mechanisms underlying endothelial senescence and their contribution to VCID are poorly understood. Inefficient DNA damage repair has been linked to vascular aging and senescence. Thus, targeting master regulators of DNA repair may increase genome stability and prevent senescence in the aged brain. E2F1 is a critical transcription factor that positively regulates the expression of genes involved in DNA repair pathways. From the existing literature, E2F1 prevents senescence in endothelial cells. We hypothesize that selective expression of E2F1 in brain endothelium can prevent endothelial senescence and improve cognitive function in aged mice. Methods: We used aged (18-months old) mice and cultured CEC derived from aged mice. We analyzed the transcriptome of cultured CEC transfected with an E2F1-encoding plasmid or an empty plasmid as control. We used an adenovirus construct to overexpress E2F1 specifically in the cerebral vasculature of aged mice, and evaluated isolated brain microvessels by immunostaining and evaluated spatial memory by behavioral tests. Results: We found that E2F1-overexpressing CEC derived from aged mice showed a reduced senescence-associated transcriptional profile, higher expression of DNA repair genes, and upregulation of tight junction gene expression, compared with control aged CEC. Micro-vessels isolated from aged mice showed hypo-phosphorylated E2F1 (reduced transcriptional activity) and enhanced p16INK4a (a tumor suppressor associated to senescence), compared with micro-vessels from young mice. Importantly, we found that selective upregulation of E2F1 in the cerebral vasculature improved spatial memory in aged mice. Conclusions: Our data suggest that selective upregulation of E2F1 in the cerebral vasculature can improve endothelial function and enhance spatial memory in aged mice.

Funding Disclosure: TARCC, American Heart Association, NIA

Poster #60

ANALYZING RELATIONSHIPS BETWEEN FATTY LIVER DISEASE, VITAMIN D, ALCOHOL CONSUMPTION, AND COGNITIVE IMPAIRMENT IN A RURAL WEST TEXAS ELDERLY COHORT

Presenting author: Nicholas Vojtkofsky, BS (Graduate Student, Texas Tech University Health Sciences Center)

Background: As the proportion of the elderly population in the U.S. continues to grow, our understanding of cognitive dysfunction becomes increasingly important. One aspect of age-dependent cognitive decline that remains relatively understudied relates to the liver-brain axis. In a preclinical model, age-dependent changes in liver function have been demonstrated hepatic trapping of vitamin A (VA) accompanied by age-dependent cognitive decline and hippocampal deficiency of the bioactive metabolite of VA, retinoic acid, in the brain. In addition, human studies have shown that non-alcoholic fatty liver disease (NAFLD) disrupts hepatic retinoid homeostasis. However, the relationship between fatty liver disease and cognition is not clear. This study aims to investigate the relationship between cognitive impairment and biomarkers of liver function as well as Vitamin D (VD) status and alcohol consumption. Methods: Cognitive function was compared to a battery of blood-based biomarkers for liver dysfunction and disease available in 299 Project FRONTIER participants (ages 62±12, 71% female, and 41% Hispanic) using Pearson correlations and linear regression analysis. Our approach examined the Repeatable Battery for Assessment of Neuropsychological Status total score (rbttlsc), serum biomarkers of liver dysfunction used in a liver function panel (ggt, ast, alt, alkp, bw_tprot, bw_tbili, trigy), serum VD (bw_vitd), and responses to an alcohol use questionnaire (aau_1). In addition, using available variables, we calculated four diagnostic indexes for NAFLD: fatty liver index (FLI), NAFLD fibrosis score (NFS), hepatic steatosis index (HSI), and the fibrosis-4 (FIB-4) index. Results: First, among individual liver biomarkers, we discovered significant negative correlations between serum bilirubin (p=0.0294) and total serum protein (p=0.0028) and RBANS total score. Second, we found significant negative correlations between FIB-4 (p=0.0108) and NFS (p=0.001) and RBANS total score. Third, we found a significant positive correlation between aau_1 and RBANS total score (p=0.0025). Finally, we discovered significant negative correlations between HSI (p<0.0001), FLI (p<0.0001), and serum alkaline phosphatase (p=0.0001) and RBANS total score. Conclusion In conclusion, our comprehensive analyses support the conclusion that liver disease is associated with a:e-related cognitive dysfunction. We identify potential diagnostic measures that could improve future protocols for identification and treatment of cognitive dysfunction in elderly populations.

Funding Disclosure: Supported by the TTUHSC Medical Student Summer Research Program (NV), TTUHSC Garrison Institute of Aging (JJL), and NIH R01 AG071859-01A1 (JJL)

Poster #61

DUAL-LANGUAGE USE AND COGNITIVE FUNCTIONING AMONG MEXICAN AMERICANS AGED 65 AND OLDER

Presenting author: Brian Downer PhD (Associate Profesor, University of Texas Medical Branch at Galveston)

Background: Bilingualism has been associated with lower dementia risk and delayed dementia onset. However, there is limited work investigating the association between individual, family, and community factors that underly differences in dual-language use and cognitive aging.This is particularly true in Hispanics, the largest population of bilingual speakers in the US. We build on prior research by investigating if bilinguals who use English and Spanish to a similar degree demonstrate better cognitive function, relative to those who use one language more than the other. Methods: We used data from waves one (1992/93) to eight (2012/13) of the Hispanic Established Population for the Epidemiological Study of the Elderly. At wave one, participants were asked what language they usually use across communicative contexts (i.e., spouse, children, family gatherings, friends, and neighbors). Degree of dual-language was based upon reported use of Spanish and English within and across contexts. Participants were categorized as low (n=1068), medium (n=696), and high (n=688) dual-language users. Cognitive functioning was measured at each observation wave with the Mini-Mental State Examination. Linear mixed models were used to estimate the association between dual-language use category, cognitive function at wave one, and change in cognition while adjusting for baseline demographic factors. Results: Participants in the medium and high dual-language use categories scored 1.47 points and 2.38 points higher at wave one compared to the low dual-language use category. Findings were stable after adjusting for language of interview. The association between dual-language use and cognition at wave one was reduced when adjusting for education (medium B=0.58 SE=0.17 p<0.01; high B=0.80 SE=0.19 p<0.01). The association between dual-language use and cognitive decline was not statistically significant in any model. Conclusion: We found that greater dual-language use is associated with higher baseline cognition but not cognitive decline among older Mexican Americans, even when accounting for differences in education level. This finding is consistent with evidence of bilingualism-induced effects on executive control, memory recall, working memory, and semantic memory. Future work should characterize bilingualism with greater nuance in order to identify the components of the bilingual experience that may serve to protect against cognitive decline in aging bilinguals.

Funding Disclosure: NIH P30AG066614, P30AG059301, P30AG024832, R01AG010939, RF1AG068988, K01AG075254

Poster #62

CHEMOTHERAPY AND RISK FOR ALZHEIMER'S DISEASE

Presenting author: Rocio Diaz Escarcega, PhD (Postdoctoral Fellow, UT Health Science Center Houston)

Background: Cognitive impairments may occur in cancer patients and survivors during or after chemotherapy. Cognitive deficits associated with neurotoxicity (chemobrain) can be subtle or disabling and frequently include disturbances in memory, attention, executive function, and processing speed. Cognitive impairments may go away soon after chemotherapy is over or may persist for years and yet, there is a paucity of effective treatments. Research has shown that chemotherapy drugs such as doxorubicin promote neurotoxicity and cognitive disturbances. Critically, some studies demonstrated that dementia occurs more commonly in cancer patients who had chemotherapy treatment compared to individuals never exposed to chemotherapy. Understanding whether and how chemotherapy may promote dementia later in life is needed. Methods: To establish new insights into chemotherapy-induced cognitive impairments, we use wild-type and Tg2576 mice (a model of Alzheimer's disease (AD)) treated with Doxil, a liposomal form of doxorubicin. Mice are injected intraperitoneally with saline or Doxil for six weeks. These conditions recapitulate a dosing schedule used in human patients and are the same as those used in similar studies in mice. Mice are then tested with cognitive and behavior assays, and their brains are analyzed for aging and AD phenotypes. Results: In our studies, we discovered that Doxil promotes cognitive impairment in wild-type mice. We also found that the brains of young mice exposed to Doxil contain lipofuscin-a mixture of oxidized proteins and lipids that is usually found only in the aged or diseased brains. DNA damage occurred with Doxil treatment in mice, confirming Doxil accelerated features of brain aging. Critically, Doxil enhances the deposition of amyloid in Tg2576 mice. Conclusion: Our study demonstrates evidence of accelerated brain aging in wild-type mice and amyloid deposition in AD mice due to Doxil treatment. Our data provide a foundation for investigating chemotherapy as a potential risk factor for AD that warrants further study. This research is being pursued in our laboratory.

Funding Disclosure: NIH R21AG067204

Poster #63

THE RELATIONSHIP BETWEEN TBI AND DEPRESSION IN MEXICAN AMERICANS WITH MILD COGNITIVE IMPAIRMENT

Presenting author: Haydee Izurieta Munoz, MS (Research Specialist, University of North Texas Health Science Center)

Background: The outcomes of Traumatic Brain Injuries (TBI) have been shown to vary depending on race and ethnicity. Studies suggest TBI as a potential risk factor for the development of dementia. The link between TBI, cognition and depression is not well understood in older Mexican Americans (MA). Our research showed that MAs with MCI and history of TBI had higher depression scores than those without. The purpose of this study is to assess depression and TBI in Mexican American elders with either amnestic or non-amnestic Mild Cognitive Impairment (MCI). Method: Participant data was obtained from the Health and Aging Brain Study: Health Disparities, a community-based research study of cognitive aging in diverse communities. Data were analyzed from 164 MAs with MCI (n=89 Non-amnestic, n=75 amnestic) and cognitive diagnosis was determined by consensus review. TBI status (obtained via self-report from Ohio State University TBI Identification method- interview form) divided participants into groups: TBI+ (at least 1 head or neck injury with loss of consciousness) and TBI- (no loss of consciousness). The Geriatric Depression Scale 30-item (GDS) was used to assess depression. Independent sample T-tests were used to determine significant differences in GDS total and subscale scores, age, education, between TBI groups (TBI + and -) split by MCI status (amnestic/non-amnestic). Results: Of n=164 MA participants with MCI, 10.4% were TBI+ for both non-amnestic and amnestic. In the non-amnestic group, there were no significant differences in age, education, total GDS or GDS subscales between TBI+ and TBI-. Amnestic MCI TBI+ participants had significantly higher total GDS scores (p=0.019) and Apathy subscale scores (p=0.019) compared to the TBI- group. There was also a trend towards significance for the subscales of Dysphoria (p=0.079) and Meaninglessness (p=0.070). There were no significant differences in age or education within the amnestic MCI participants. Conclusion: Older MAs with amnestic MCI with TBI+ had significantly increased total depression, as well as Apathy. Although the role of TBI in mood disorders is not fully understood, additional research in this area could provide insight into the importance of cognitive status when evaluating depression.

Funding Disclosure: National Institute on Aging Award Numbers R01AG054073 and R56AG058533

Poster #64

IMPROVING CARE TRANSITIONS FOR CAREGIVERS OF HOSPITALIZED VETERANS LIVING WITH DEMENTIA: A QUALITATIVE STUDY OF HEALTH PROFESSIONALS

Presenting author: Molly Horstman, MD, MS (Assistant Professor, Baylor College of Medicine)

Background: Hospital admissions are stressful for adults with dementia and their caregivers. During care transitions from hospital to home, outcomes for adults with dementia depend, in part, on the caregiver's health and well-being. We aimed to identify the resources and training needs of dementia caregivers during care transitions. Methods: We conducted semi-structured interviews with licensed health professionals. Eligible health professionals were those who provide care for Veterans with Alzheimer's Disease and Related Dementias (ADRD) at the Michael E. DeBakey Veterans Affairs (VA) Medical Center in Houston, TX. Interviews were recorded and transcribed verbatim. We conducted a rapid qualitative analysis with structured summaries and matrices. Results: Fifteen health professionals completed a semi-structured interview (4 Social Workers, 4 Hospitalists, 2 Inpatient Nurse Case Managers, 2 Geriatrics Providers, 1 Primary Care Provider, and 2 Geriatric Psychiatrists). Health professionals identified opportunities to improve the information provided to dementia caregivers in the hospital with an emphasis on supporting caregivers to set "realistic expectations" for dementia progression and the availability of VA and community-based services and supports. Although training is determined on a "case-by-case basis", six health professionals recommended that caregiver training on behavioral interventions to address dementia behaviors should be available in the hospital. Interviewees perceived that services that provided an "extra pair of hands" were most beneficial to caregivers after discharge, such as home health aides. Across health professions, interviewees recommended that caregivers receive clear instructions on who to contact following discharge if problems arise. Interviewees recommended a dementia "navigator for the caregivers" who would meet the caregiver in the hospital and follow up with the caregiver after discharge to assist with post-discharge problem-solving, provide dementia-specific education, and connect the caregiver to VA and non-VA services and supports. Conclusions: Health professionals identified several opportunities to improve the hospital and care transition experience for caregivers of hospitalized veterans with dementia. Adding a caregiver navigator who can follow the caregiver from the hospital to home was recommended across health professions and is supported by evidence-based care transitions interventions.

Funding Disclosure: VA HSR&D CDA, National Institute on Aging

Poster #65

THE INTERPLAY BETWEEN ALL-TRANS RETINOIC ACID, THE GUT-BRAIN AXIS, AND THE PATHOGENESIS OF ALZHEIMER'S DISEASE: A POTENTIAL MECHANISM

Presenting author: Matthew Buxton, BS (Medical Student, Texas Tech University Health Sciences Center)

Background: Alzheimer's disease (AD) is a debilitating disease affecting 6.5 million people in the United States (US) presently, anticipating to impact approximately 12.7 million people in the US by 2050. Recent studies demonstrating bacterial lipopolysaccharide (LPS) is present in the hippocampus of AD patients has warranted a closer examination of the role of the gastrointestinal (GI) tract integrity and the gut-brain axis (GBA) in AD. There is growing evidence that microbiota composition of the GI tract influences the cascade of signals sent throughout the complex system of the GBA. In relation to AD, the appearance of LPS in the post-mortem hippocampus could signify a compromised mucosal barrier and tight junction coupling between GI endothelial cells, which subsequently alters the GBA. The breakdown of mucosal and epithelial barrier integrity, combined with immune system activation, results in a pro-pathologic state that creates gut microbial dysbiosis. We propose tight junction deficiencies found in pathologic states arise primarily through preventable environmental causes. Methods: A standard literature search including PubMed was used to identify knowledge gaps and molecular interrelationships between VA and AD. Results: There is evidence of Vitamin A (VA) dysregulation in AD pathogenesis and progression. Our currently funded NIH R01 grant addresses how brain transcriptomic, metabolomic and lipidomic profiles are altered by VA deficiency. All-trans retinoic acid (ATRA) is essential for normal expression of the tight junction proteins ZO-1, occludin, and claudin-1. VA deficiency leads to decreased expression of these proteins and subsequently compromises tight junctions. VA is also required for the differentiation of Treg cells. We propose VA deficiency compromises tight junctions, which results in pathologic metabolites like LPS penetrating the epithelium and entering the bloodstream. The loss of Treg cells cannot counteract the proinflammatory response, beginning an escalating cycle of inflammation. In addition, butyrate, a short-chain fatty acid (SCFA) produced from Firmicutes through fiber fermentation, produces retinoic acid in gut epithelial cells by inhibiting histone deacetylases. Conclusion: The proposed mechanism attempts to explain the complex interrelationship between VA, butyrate, AD, and the GBA. Pharmacologic treatments targeted at such a mechanism could ameliorate and decrease the risk of developing Alzheimer's Disease.

Funding Disclosure: Supported by the Medical Student Summer Research Program (MB) and NIH R01AG071859.

Poster #66

COGNITIVE AND SLEEP DIFFERENCES IN INDIVIDUALS WITH PARKINSON'S DISEASE AND RAPID EYE MOVEMENT SLEEP BEHAVIOR DISORDER

Presenting author: Vanessa M. Young, MS (Clinical Research Coordinator, UT Health Science Center San Antonio)

Background: Rapid eye movement (REM) sleep behavior disorder (RBD) is a parasomnia characterized by dream reenactment and simple or complex abnormal motor behaviors during the REM sleep stage. RBD can be an early predictor of Parkinson's disease (PD) and dementia; however, the relationship between dementia, RBD, and broader sleep disturbance in PD remains unclear. The aim of this pilot study was to objectively assess sleep differences among individuals with PD with and without RBD using polysomnography. Method: Twenty adults with PD were enrolled in the study. The presence of RBD was assessed with the validated Mayo Sleep Questionnaire. Participants completed neuropsychological evaluations, physical functioning assessments, questionnaires, digital device sleep monitoring, and an overnight in-clinic polysomnography. Demographic and objective sleep differences between the RBD and non-RBD groups were evaluated using independent sample t-tests for continuous variables and chi-square tests for categorical variables. Results: The RBD group (n=12) and the non-RBD group (n=8) showed no statistically significant differences in educational level, age, and sex distribution (all p>0.05). However, the RBD group displayed poorer global cognition as assessed by the Montreal Cognitive Assessment (RBD mean = 23.73 ± 4.29, non-RBD mean = 26.88 ± 2.23, t(17)= -3.11, p=0.006). On polysomnography, the RBD group trended towards a lower average sleep duration (t(18)=-2.0, p=0.06), and displayed significantly lower percentage of REM sleep (3.40 ± 1.68; t(18)=-2.6, p=0.02). There were no group differences in percentages of Stage 1 and 2 sleep and the number of periodic limb movements (all p>0.05). Conclusion: Individuals with PD and RBD displayed poorer global cognition relative to individuals with PD without RBD. On the polysomnography, individuals with PD and RBD also trended towards lower total sleep time and less time spent in REM sleep. The results suggest broad sleep differences in RBD and non-RBD groups, which have been associated with increased dementia risk in other samples. However, the study was underpowered to fully explore these differences and the one-night sleep assessment may not be representative of typical sleep at home hindering generalization. Future studies incorporating digital sleep monitoring devices are underway.

Funding Disclosure: UTHSCSA Perry & Ruby Stevens Foundation Parkinson's Disease Center of Excellence and NIH P30AG066546