INS NYC 2024 Program

Poster	Poster Session 02 Program Schedule 02/15/2024 08:00 am - 09:15 am Room: Shubert Complex (Posters 1-60) Poster Session 02: Aging \| MCI \| Neurodegenerative Disease - PART 1 Final Abstract #4 Serial Position Effect Profiles and Their Neuroanatomical Correlates in Healthy Controls and Mild Cognitive Impairment Isabelle Avildsen, New York University Langone Medical Center, New York, United States Nancy Foldi, Weill Cornell Medical School & Queens College, City University of New York, New York, United States Aditya Kulkarni, Massachusetts General Hospital, Harvard Medical School, Cambridge, United States Melissa Pocsai, Queens College & The Graduate Center, City University of New York, New York, United States Vivian Chu, Queens College & The Graduate Center, City University of New York, New York, United States Category: Dementia (Alzheimer's Disease) Keyword 1: neuroimaging: structural Keyword 2: semantic processing Keyword 3: mild cognitive impairment Objective: This study investigated how serial position effect (SPE) measures relate to risk of Alzheimer’s disease (AD) in a cohort of healthy controls (HC) and individuals diagnosed with mild cognitive impairment (MCI). AD-related disruptions to semantic and mnemonic networks are hypothesized to create specific learning and recall patterns from the primacy, middle and recency list positions. Previous work identifies poor primacy item recall as particularly sensitive to conversion risk to MCI/AD. Thus, we hypothesized that sensitive SPE metrics would be subserved by distinct neuroanatomical correlates associated with differential levels of encoding. Specifically, primacy measures would associate with areas of semantic processing (hippocampal, medial temporal, and/or frontal lobe regions), while recency scores would associate with areas of non-semantic, shallow processing of language. Participants and Methods: Participants were individuals diagnosed as either HC (N = 158) or MCI (N = 269) from the Alzheimer’s Disease Neuroimaging Initiative. Scores from the Rey Auditory Verbal Learning Test (RAVLT) at Learning, Short Delay (SD), and Long Delay (LD) trials were derived as follows: percent accuracy recall for each SPE (Primacy, Middle, Recency) and total recall scores (Total-RAVLT). Multivariable linear regression analyses were performed for each SPE and the Total-RAVLT score. Analyses controlled for age, education, sex, race, APOE-ε4 allele status, and conversion status. Data were stratified by baseline diagnosis and hemisphere, and thereby examined the association between regional brain volumes, cortical thickness, and SPE scores. To reduce Type 1 error rate from multiple comparisons, alpha was set to .001. Results: Neuroanatomical correlates of SPE scores emerged as follows. HC: Only Recency at LD was associated with left medial orbitofrontal cortical thickness, and no Total-RAVLT score was significantly associated with any neuroanatomical correlate. MCI: Primacy and Middle scores across trials were associated with varied bilateral and unilateral regional volumes and thicknesses (e.g., superior parietal, medial orbitofrontal, pars orbitalis, fusiform, superior frontal, precuneus, rostral middle frontal, inferior temporal, insula), as were Total-RAVLT scores. Conclusions: The current study sought to elucidate specific neuroanatomical correlates of SPE metrics given prior evidence that serial position profiles are sensitive to risk of MCI/AD. In HC, only recency at LD correlated with cortical thickness, while no Total-RAVLT scores were associated with neuroanatomical regions. In MCI, Primacy scores were associated with multiple cortical volumes and thicknesses, which, as hypothesized, subserve known regions of semantic and mnemonic processing. Together, SPE markers are sensitive to preclinical neuroanatomical changes – at times even more so than total scores – where Primacy position accuracy in MCI are associated with AD-related neuroanatomical changes. We conclude that cognitive markers of SPE, which draw on semantic and mnemonic functions, can improve detection of future decline given their robust associations with regions known to degrade in AD.

Poster

02/15/2024
08:00 am - 09:15 am
Room: Shubert Complex (Posters 1-60)

Poster Session 02: Aging | MCI | Neurodegenerative Disease - PART 1

Final Abstract #4

Serial Position Effect Profiles and Their Neuroanatomical Correlates in Healthy Controls and Mild Cognitive Impairment

Isabelle Avildsen, New York University Langone Medical Center, New York, United States
Nancy Foldi, Weill Cornell Medical School & Queens College, City University of New York, New York, United States
Aditya Kulkarni, Massachusetts General Hospital, Harvard Medical School, Cambridge, United States
Melissa Pocsai, Queens College & The Graduate Center, City University of New York, New York, United States
Vivian Chu, Queens College & The Graduate Center, City University of New York, New York, United States

Category: Dementia (Alzheimer's Disease)

Keyword 1: neuroimaging: structural
Keyword 2: semantic processing
Keyword 3: mild cognitive impairment

Objective:

This study investigated how serial position effect (SPE) measures relate to risk of Alzheimer’s disease (AD) in a cohort of healthy controls (HC) and individuals diagnosed with mild cognitive impairment (MCI). AD-related disruptions to semantic and mnemonic networks are hypothesized to create specific learning and recall patterns from the primacy, middle and recency list positions. Previous work identifies poor primacy item recall as particularly sensitive to conversion risk to MCI/AD. Thus, we hypothesized that sensitive SPE metrics would be subserved by distinct neuroanatomical correlates associated with differential levels of encoding. Specifically, primacy measures would associate with areas of semantic processing (hippocampal, medial temporal, and/or frontal lobe regions), while recency scores would associate with areas of non-semantic, shallow processing of language.

Participants and Methods:

Participants were individuals diagnosed as either HC (N = 158) or MCI (N = 269) from the Alzheimer’s Disease Neuroimaging Initiative. Scores from the Rey Auditory Verbal Learning Test (RAVLT) at Learning, Short Delay (SD), and Long Delay (LD) trials were derived as follows: percent accuracy recall for each SPE (Primacy, Middle, Recency) and total recall scores (Total-RAVLT). Multivariable linear regression analyses were performed for each SPE and the Total-RAVLT score. Analyses controlled for age, education, sex, race, APOE-ε4 allele status, and conversion status. Data were stratified by baseline diagnosis and hemisphere, and thereby examined the association between regional brain volumes, cortical thickness, and SPE scores. To reduce Type 1 error rate from multiple comparisons, alpha was set to .001.

Results:

Neuroanatomical correlates of SPE scores emerged as follows. HC: Only Recency at LD was associated with left medial orbitofrontal cortical thickness, and no Total-RAVLT score was significantly associated with any neuroanatomical correlate. MCI: Primacy and Middle scores across trials were associated with varied bilateral and unilateral regional volumes and thicknesses (e.g., superior parietal, medial orbitofrontal, pars orbitalis, fusiform, superior frontal, precuneus, rostral middle frontal, inferior temporal, insula), as were Total-RAVLT scores.

Conclusions:

The current study sought to elucidate specific neuroanatomical correlates of SPE metrics given prior evidence that serial position profiles are sensitive to risk of MCI/AD. In HC, only recency at LD correlated with cortical thickness, while no Total-RAVLT scores were associated with neuroanatomical regions. In MCI, Primacy scores were associated with multiple cortical volumes and thicknesses, which, as hypothesized, subserve known regions of semantic and mnemonic processing. Together, SPE markers are sensitive to preclinical neuroanatomical changes – at times even more so than total scores – where Primacy position accuracy in MCI are associated with AD-related neuroanatomical changes. We conclude that cognitive markers of SPE, which draw on semantic and mnemonic functions, can improve detection of future decline given their robust associations with regions known to degrade in AD.