INS NYC 2024 Program

Poster	Poster Session 02 Program Schedule 02/15/2024 08:00 am - 09:15 am Room: Majestic Complex (Posters 61-120) Poster Session 02: Aging \| MCI \| Neurodegenerative Disease - PART 1 Final Abstract #84 Verbal Memory Algorithm Predicts Conversion to Dementia in a Clinical Sample with MCI Victoria Sanborn, Alpert Medical School at Brown University/Rhode Island Hospital - Lifespan, Providence, United States Aimee Karstens, Mayo Clinic; Alpert Medical School at Brown University, Rochester, United States Zachary Kunicki, Alpert Medical School at Brown University, Providence, United States Sarah Hegedus, Providence College, Providence, United States Melissa Zammitti, Rhode Island Hospital-Lifespan; Pacific University, Forest Grove, United States Geoffrey Tremont, Alpert Medical School at Brown University/Rhode Island Hospital - Lifespan, Providence, United States Category: Dementia (Alzheimer's Disease) Keyword 1: assessment Keyword 2: mild cognitive impairment Keyword 3: learning Objective: Mild cognitive impairment (MCI) subtypes characterize research samples to determine risk of developing Alzheimer’s disease (AD) and other neurocognitive disorders. The utility of these subtypes is less evident in clinical samples, and other cognitive characteristics (e.g., poor encoding versus retrieval versus spontaneous recall) are often used to make real-world diagnoses. We examined the usefulness of a verbal memory algorithm for predicting conversion to dementia in a sample of older adult outpatients with MCI. Participants and Methods: Archival data from 60 older adult patients diagnosed with MCI through an outpatient neuropsychology clinic who completed a neuropsychological evaluation (Eval1) and re-evaluation (Eval2) were included. Patients were classified into subgroups based on a verbal memory performance algorithm using scores from the Hopkins Verbal Learning Test-Revised (HVLT). Groups were delineated based on yes/no classification of 1) Impaired learning then 2) Impaired recall. Age-corrected scores ≥1.5 standard deviations (SD) below published normative values were considered impaired. Subgroups included: 1) Learning intact/memory intact (Intact), 2) Learning impaired/memory intact (Learning), 3) Learning intact/memory impaired (Memory), 4) Learning impaired/memory impaired (Combined). Information from neuropsychological reports were coded based on McKhann et al. (2015) criteria to classify conversion to dementia from Eval1 to Eval2. Pearson’s chi-square analysis and one-way ANOVAs compared subgroups on demographic variables and global cognitive function (Dementia Rating Scale-2 Scaled Score; DRS) at Eval1. Logistic regressions controlling for age, education, sex, number of years from Eval1 to Eval2, and global impairment (DRS) examined risk of conversion among the subgroups. Results: The verbal memory algorithm captured all 60 patients. The Learning group (n=3) and Intact group (n=14) were merged to promote equality of group sizes (n=17; Memory n=17, Combined n=26). There were no significant differences between groups for age, education, sex, or time from Eval1 to Eval2 (all p’s >0.05). Average number of years from Eval1 to Eval2 was 2.28 (SD=1.29). The Intact group showed the highest baseline global cognitive performance (DRS M=8.9, SD=2.1) followed by the Memory group (DRS M=8.3, SD=2.6) then the Combined group (DRS M=7.3, SD=2.1). The Combined group had over 6 times greater risk for conversion to dementia from Eval1 to Eval2 than the Intact group (OR=6.6, p=0.03) and the Memory group (OR=6.6, p=0.03) when controlling for age, education, years from Eval1 to Eval 2, and DRS. Conclusions: A verbal memory algorithm showed possible utility in identifying patients with elevated risk for converting from MCI to dementia in a clinical sample, even when controlling for global cognitive impairment. Patients with an amnestic-only pattern of verbal memory performance showed similar rates of conversion to dementia as patients without memory impairment, while those with impaired learning and memory were at greatest risk for conversion within 2-3 years of MCI diagnosis. Future research should integrate additional tests of memory and language to determine if this algorithm holds and examine its utility for early detection of specific etiologies, such as Alzheimer’s disease.

Poster

02/15/2024
08:00 am - 09:15 am
Room: Majestic Complex (Posters 61-120)

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

Final Abstract #84

Verbal Memory Algorithm Predicts Conversion to Dementia in a Clinical Sample with MCI

Victoria Sanborn, Alpert Medical School at Brown University/Rhode Island Hospital - Lifespan, Providence, United States
Aimee Karstens, Mayo Clinic; Alpert Medical School at Brown University, Rochester, United States
Zachary Kunicki, Alpert Medical School at Brown University, Providence, United States
Sarah Hegedus, Providence College, Providence, United States
Melissa Zammitti, Rhode Island Hospital-Lifespan; Pacific University, Forest Grove, United States
Geoffrey Tremont, Alpert Medical School at Brown University/Rhode Island Hospital - Lifespan, Providence, United States

Category: Dementia (Alzheimer's Disease)

Keyword 1: assessment
Keyword 2: mild cognitive impairment
Keyword 3: learning

Objective:

Mild cognitive impairment (MCI) subtypes characterize research samples to determine risk of developing Alzheimer’s disease (AD) and other neurocognitive disorders. The utility of these subtypes is less evident in clinical samples, and other cognitive characteristics (e.g., poor encoding versus retrieval versus spontaneous recall) are often used to make real-world diagnoses. We examined the usefulness of a verbal memory algorithm for predicting conversion to dementia in a sample of older adult outpatients with MCI.

Participants and Methods:

Archival data from 60 older adult patients diagnosed with MCI through an outpatient neuropsychology clinic who completed a neuropsychological evaluation (Eval1) and re-evaluation (Eval2) were included. Patients were classified into subgroups based on a verbal memory performance algorithm using scores from the Hopkins Verbal Learning Test-Revised (HVLT). Groups were delineated based on yes/no classification of 1) Impaired learning then 2) Impaired recall. Age-corrected scores ≥1.5 standard deviations (SD) below published normative values were considered impaired. Subgroups included: 1) Learning intact/memory intact (Intact), 2) Learning impaired/memory intact (Learning), 3) Learning intact/memory impaired (Memory), 4) Learning impaired/memory impaired (Combined). Information from neuropsychological reports were coded based on McKhann et al. (2015) criteria to classify conversion to dementia from Eval1 to Eval2. Pearson’s chi-square analysis and one-way ANOVAs compared subgroups on demographic variables and global cognitive function (Dementia Rating Scale-2 Scaled Score; DRS) at Eval1. Logistic regressions controlling for age, education, sex, number of years from Eval1 to Eval2, and global impairment (DRS) examined risk of conversion among the subgroups.

Results:

The verbal memory algorithm captured all 60 patients. The Learning group (n=3) and Intact group (n=14) were merged to promote equality of group sizes (n=17; Memory n=17, Combined n=26). There were no significant differences between groups for age, education, sex, or time from Eval1 to Eval2 (all p’s >0.05). Average number of years from Eval1 to Eval2 was 2.28 (SD=1.29). The Intact group showed the highest baseline global cognitive performance (DRS M=8.9, SD=2.1) followed by the Memory group (DRS M=8.3, SD=2.6) then the Combined group (DRS M=7.3, SD=2.1). The Combined group had over 6 times greater risk for conversion to dementia from Eval1 to Eval2 than the Intact group (OR=6.6, p=0.03) and the Memory group (OR=6.6, p=0.03) when controlling for age, education, years from Eval1 to Eval 2, and DRS.

Conclusions:

A verbal memory algorithm showed possible utility in identifying patients with elevated risk for converting from MCI to dementia in a clinical sample, even when controlling for global cognitive impairment. Patients with an amnestic-only pattern of verbal memory performance showed similar rates of conversion to dementia as patients without memory impairment, while those with impaired learning and memory were at greatest risk for conversion within 2-3 years of MCI diagnosis. Future research should integrate additional tests of memory and language to determine if this algorithm holds and examine its utility for early detection of specific etiologies, such as Alzheimer’s disease.