Back to Search Start Over

Cognitive Clusters of Early‐onset non‐Alzheimer's disease.

Authors :
Kim, Hee Jin
Lane, Kathleen A.
Gao, Sujuan
Hammers, Dustin B.
Logan, Paige E.
Aisen, Paul S. S
Dage, Jeff L.
Eloyan, Ani
Foroud, Tatiana M.
Grinberg, Lea T.
Iaccarino, Leo
Jack, Clifford R.
Kramer, Joel H.
Koeppe, Robert A.
Kukull, Walter A.
La Joie, Renaud
Mundada, Nidhi S
Murray, Melissa E.
Nudelman, Kelly N. H.
Rumbaugh, Malia C.
Source :
Alzheimer's & Dementia: The Journal of the Alzheimer's Association; Dec2023 Supplement 24, Vol. 19, p1-4, 4p
Publication Year :
2023

Abstract

Background: Approximately 10‐25% of amnestic early‐onset dementia are negative for amyloid (Early‐onset non‐Alzheimer's Disease, EOnonAD). EOnonAD is a heterogenous group with various etiologies. In order to better understand the traits of EOnonAD, we clustered amnestic EOnonAD individuals according to cognitive patterns. Method: Cognitive data of 62 EOnonAD individuals from the Longitudinal Early‐onset AD study (LEADS) were Z‐normalized to 91 intact individuals after adjusting for age, sex, and education. This led to domain‐averaged composite Z‐scores for memory, language, speed/attention, visuospatial, and executive function. Cluster analysis using the Ward method on the residuals for the five composites was performed. We used Scree plot using the pseudo T‐squared to determine the optimal number of clusters. We compared demographics, imaging biomarkers (amyloid and tau PET SUVR, hippocampal volume), plasma (pTau231, Aß40, Aß42, glial fibrillary acidic protein, and neurofilament light chain (NfL)), and cerebrospinal fluid (CSF) biomarkers (Aß42/40, NfL, Neurogranin, Visinin‐like protein, YKL‐40, and Synaptosome Associated Protein 25) across the clusters after adjusting for age, sex, and education. We compared gray matter density of each EOnonAD cluster to controls (n = 84) using voxel‐wise multiple linear regressions in SPM12, adjusting for age, sex, education, and intracranial volume. Result: Scree plot using the pseudo‐T‐squared suggested three clusters. Cluster 1 (38/62, 61.3%) was characterized by having the mildest cognitive impairment in all domains. Cluster 2 (19/62, 30.6%) was characterized by more severe memory and language impairment. Cluster 3 (5/62, 8.1%) was characterized by more severe impairment in speed/attention, visuospatial, and executive function (Table 1). Cluster 3 showed significantly higher NfL for both plasma (Cluster 1 vs 3, p<0.001; Cluster 2 vs 3, p < 0.004) and CSF (Cluster 1 vs 3, p = 0.001). Other biomarkers did not show significant difference (Table 1). Only Cluster 3 showed significant atrophy in the bilateral fronto‐parietal areas at p<0.01, uncorrected (Figure 1). Conclusion: We identified heterogeneity in cognitive patterns among amnestic EOnonAD individuals. Cluster 1 had mildest cognitive impairment; Cluster 2 had predominant impairment in memory and language; and Cluster 3 had predominant impairment in speed/attention, visuospatial, and executive function, with significant fronto‐parietal atrophy and high NfL indicating active neurodegeneration. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
15525260
Volume :
19
Database :
Supplemental Index
Journal :
Alzheimer's & Dementia: The Journal of the Alzheimer's Association
Publication Type :
Academic Journal
Accession number :
174412061
Full Text :
https://doi.org/10.1002/alz.083061