Free Recall Outperforms Story Recall in Associations with Plasma Biomarkers in Preclinical Alzheimer Disease

Abstract

Background

A decline in episodic memory is one of the earliest cognitive characteristics of Alzheimer disease and memory tests are heavily featured in cognitive composite endpoints that are used to demonstrate treatment efficacy. Assessments of episodic memory can take many forms including free recall, associate learning, and paragraph or story recall. Plasma biomarkers of Alzheimer disease are now widely available and will likely form the backbone of cohort enrichment strategies for future clinical trials. Thus, it is critical to evaluate which episodic memory measures are most sensitive to plasma markers of Alzheimer disease pathology.

Objectives

To compare the associations of common episodic memory tests with plasma biomarkers of Alzheimer disease.

Design

Longitudinal cohort study.

Setting

Academic medical center in the midwestern United States.

Participants

A total of 161 cognitively normal older adults with at least one plasma biomarker assessment and two or more annual clinical and cognitive assessments which included up to three different tests of episodic memory.

Measurements

Episodic memory performance using free recall, paired associates recall or paragraph recall. Plasma Aβ42, Aβ40, ptau217, and neurofilament light chain were measured.

Results

Free recall on the Free and Cued Selective Reminding Test with Immediate Recall (FCSRT + IR) was substantially more sensitive to longitudinal cognitive change associated with abnormal baseline plasma Aβ42/Aβ40 and ptau217 compared to other measures of episodic memory. A cognitive composite that included only free recall showed larger decline associated with baseline Aβ42/Aβ40 when compared to those that included paragraph recall. Differences in decline across composites were minimal when considering baseline ptau217 or NfL.

Conclusion

Episodic memory is a critical domain to assess in preclinical Alzheimer disease. Methods of assessing memory are not equal and longitudinal change in free recall substantially outperformed both paired associates and paragraph recall. Clinical trial results will depend critically on the episodic memory test(s) that are chosen for a composite endpoint and free recall from the FCSRT + IR is an optimal memory measure to include rather than paired associates or paragraph recall.