4,068 posters, 219 videos, 101 topics, 69 sessions, 14,365 authors, 6,128 institutions

Alzheimer's Association International Conference 2026

July 12-15, 2026 | London, United Kingdom, and Online

8940

Domain-specific Cognitive Profiling From Informant-based Assessments In A Population-based Sample

View Poster Contact Presenter

Poster Presenter

Raul Ururahy

Authors

Raul Ururahy, Laiss Bertola, Renata Leite, Lea Grinberg, Eduardo Ferriolli, Claudia Suemoto

Affiliations

Division of Geriatrics, University of São Paulo Medical School, São Paulo, Brazil, Laboratory Medicine and Pathology, Mayo Clinic, Jacksonville, Florida, United States of America

Part of Topic

Introduction:

COGNITIVE ASSESSMENT IN STUDIES

- Clinicopathological studies in population-based cohorts offer the most reliable framework for linking cognitive phenotypes to neuropathology.1

Complexity and cost often limit sample size and

the depth of cognitive characterization.2

- Many studies rely on global cognitive outcomes, which insufficiently capture domain-specific patterns of impairment.1

- Widely used instruments such as the Clinical Dementia Rating (CDR) and the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) include items that implicitly assess multiple cognitive domains.3

- Unlocking this embedded information provides an opportunity to derive more granular cognitive information without additional testing or cost.

Objective:

-Extract domain-specific cognitive constructs from global cognition instruments using informant-based data.

Methods:

-Population-based Brazilian post-mortem sample (Biobank for Aging Studies) (n=4,504)

-Cognitive abilities assessed using the informant-based version of the CDR scale and the IQCODE.

> Global score and CDR sum of boxes (CDR-SOB)

> IQCODE 26-item version

-Informant-based assessments of cognitive and functional status in post-mortem settings have been shown to provide reliable and valid estimates for dementia diagnosis.

-Dataset constructed comprising the full interview from the informant section of the CDR and the 26 items of the IQCODE.

-Suitability of the sample assessed using the Kaiser–Meyer–Olkin (KMO) measure, and Bartlett’s test of sphericity.

-Exploratory factor analysis conducted using principal axis factoring with oblique rotation.

-Factor retention guided by parallel analysis and scree plot inspection. Item performance, communalities, and cross-loadings were systematically evaluated through iterative model refinement.

Results:

- Sample characteristics:

> Mean age=71.2±13.4 yo; 46.6% women; 62.9% White

> 74.4% CDR=0 ; mean IQCODE=3.3±0.6

- Overall KMO=0.99 and Bartlett’s test p<0.001 (excellent dataset factorability)

- Iterative analyses supported a stable three-factor structure

> Factor 1 (disproportionately large share of common variance): Items spanning multiple cognitive domains, consistent with a dominant general cognitive dimension.

> Factors 2 and 3 (smaller contribution to common variance): Primarily functional and self-care items.

- Factor 1 prioritized as the foundation for a cognitive domain assessment tool, given its conceptual breadth and statistical predominance.

- Items from Factor 1 assigned to specific cognitive domains using theory-driven considerations.

> Currently undergoing expert panel content validation, with confirmatory factor analysis planned.

Conclusion:

- Domain-specific cognitive information can be derived from global cognitive instruments, including informant-based assessments.

References:

1. Cholerton B et al (J Alzheimer’s Dis. 2013)

2. Ong KT (Challenges in Dementia Studies. 2018)

3. Burton L et al (J Rehabil Med. 2015)

4. Ferretti REDL et al (Dement Neuropsychol. 2010)

Comments

Email is required but is not shown publicly.

Sort comments

Loading comments...