Impact of global amyloid SUVR on Spanish-English Verbal Learning Test (SEVLT) by race and ethnicity
Poster No:
915
Submission Type:
Abstract Submission
Authors:
Rajesh Nandy1, Sid O'Bryant1, Melissa Petersen1, James Hall1
Institutions:
1University of North Texas Health Science Center, Fort Worth, TX
First Author:
Co-Author(s):
Introduction:
Association of global amyloid beta (Aβ) burden with cognitive performance has been studied extensively. Most studies have found a statistically significant negative correlation between global amyloid beta (Aβ) burden in the brain and cognitive scores, indicating a potential link to cognitive decline and Alzheimer's disease. Many studies also considered other covariates like sex, APOE genotypes etc. However, very few studies have performed a comprehensive investigation on the effect of race/ethnicity on the association. In this study, we use Spanish- English Verbal Learning Test (SEVLT) scores (validated for Hispanic and non-Hispanic populations) as the primary measure of memory performance and global standardized uptake value ratio (SUVR) as a measure of amyloid burden for a large multi-ethnic cohort of aging population.
Methods:
Data - The data used here are obtained from Release 5 of the Health and Aging Brain Study - Health Disparities (HABS-HD) study. A total of 2125 participants are included with the following demographic breakdown – 642 Hispanics (H), 729 non-Hispanic Blacks (NHB) and 754 non-Hispanic Whites (NHW); 1354 Females and 771 males; Age ranging from 50-95. Spanish English Verbal Learning Test (SEVLT), a verbal episodic learning and memory test that was developed as a valid and comparable measure for use in both English and Spanish. The MMSE was used as a measure of global cognition.
Statistical Analysis – Multiple regression model was used with cognitive performance score as the outcome with the following covariates – Global SUVR, Ethnicity (H, NHW & NHB), Age, Years of education, Gender. We also considered potential interaction-effects of ethnicity with the other covariates. For Ethnicity, Hispanics were considered to be the reference group .
Statistical Analysis – Multiple regression model was used with cognitive performance score as the outcome with the following covariates – Global SUVR, Ethnicity (H, NHW & NHB), Age, Years of education, Gender. We also considered potential interaction-effects of ethnicity with the other covariates. For Ethnicity, Hispanics were considered to be the reference group .
Results:
In Figure 1, we present the results with SELVT as the outcome. All the main effects are highly significant. For unit increase in Global SUVR, SELVT is expected to decrease by 8.2. NHB group does not have a significant difference with the H group, but NHW group on average scores higher by 11.03. Females, on average score higher by 4.43. The score decreases by 0.13, on average, for each year increase in age. Each year of education increases the score by 0.73. Only NHW group interacted significantly with Age, Education and Gender. They exhibited higher decline by age, lower increase in performance by education and higher performance difference among genders.
In Figure 2, we present the results with SELVT-DR (delayed recall) as the outcome. This time neither Ethnicity, nor any of the interactions are significant. All the other main effects are highly significant. For unit increase in Global SUVR, SELVT-DR is expected to decrease by 4.3. Females, on average score higher by 1.94. The score decreases by 0.08, on average, for each year increase in age. Each year of education increases the score by 0.24. It appears that all the effects of Global SUVR are stronger for delayed recall, as evidenced by a p-value close to zero. However, the statistically significant differences are directionally the same, as expected.
Finally, we also ran similar analysis using MMSE, the more commonly used measure of global cognition. The effects were much smaller compared to SELVT and SELVT-DR. However, it demonstrated significant interaction between Global SUVR and ethnicity with a stronger decline in performance with increased SUVR for the NHB and NHW groups.
In Figure 2, we present the results with SELVT-DR (delayed recall) as the outcome. This time neither Ethnicity, nor any of the interactions are significant. All the other main effects are highly significant. For unit increase in Global SUVR, SELVT-DR is expected to decrease by 4.3. Females, on average score higher by 1.94. The score decreases by 0.08, on average, for each year increase in age. Each year of education increases the score by 0.24. It appears that all the effects of Global SUVR are stronger for delayed recall, as evidenced by a p-value close to zero. However, the statistically significant differences are directionally the same, as expected.
Finally, we also ran similar analysis using MMSE, the more commonly used measure of global cognition. The effects were much smaller compared to SELVT and SELVT-DR. However, it demonstrated significant interaction between Global SUVR and ethnicity with a stronger decline in performance with increased SUVR for the NHB and NHW groups.
·Figure 2. Results for SEVLT-DR (delayed recall)
·Figure 1. Results for SEVLT
Conclusions:
This is the largest examination of the link between cerebral amyloid burden and cognitive outcomes in a multi-ethnic study to date. This study demonstrates that global cerebral amyloid burden is more strongly associated with delayed recall as compared to immediate memory as compared to global cognition. The link between global amyloid burden and global cognition was stronger among NHBs. Future work will examine the impact of APOE4 genotype and specific amyloid regions of interest on these outcomes.
Lifespan Development:
Aging 1
Modeling and Analysis Methods:
Classification and Predictive Modeling 2
Keywords:
Aging
Cognition
Data analysis
Positron Emission Tomography (PET)
1|2Indicates the priority used for review