Face-name associative memory performance is related to amyloid burden in normal elderly

Abstract

Cerebral amyloid beta (Aβ) deposition occurs in a substantial fraction of cognitively normal (CN) older individuals. However, it has been difficult to reliably detect evidence of amyloid-related cognitive alterations in CN using standard neuropsychological measures. We sought to determine whether a highly demanding face-name associative memory exam (FNAME) could detect evidence of Aβ-related memory impairment in CN. We studied 45 CN subjects (mean age=71.7 ± 8.8) with Clinical Dementia Rating (CDR) scores=0 and MMSE ≥ 28, using Positron Emission Tomography with Pittsburgh Compound B (PiB PET). Memory factor scores were derived from a principal components analysis for FNAME name retrieval (FN-N), FNAME occupation retrieval (FN-O) and the 6-Trial Selective Reminding Test (SRT). Using multiple linear and logistic regression analyses, we related the memory factor scores to PiB distribution volume ratios (DVR, cerebellar reference) as either a continuous or a dichotomous variable in frontal cortex and a posterior cortical region representing the precuneus, posterior cingulate and lateral parietal cortices (PPCLP), co-varying for age and AMNART IQ (a proxy of cognitive reserve (CR)). A significant inverse relationship for FN-N was found with Aβ deposition in frontal (R(2)=0.29, β=-2.2, p=0.02) and PPCLP cortices (R(2)=0.26, β=-2.4, p=0.05). In contrast, neither FN-O nor the SRT were significantly related to Aβ deposition. Performance on a demanding test of face-name associative memory was related to Aβ burden in brain regions associated with memory systems. Associative memory for faces and names, a common complaint among older adults, may be a sensitive marker of early Aβ-related impairment.

Figure 1

Figure 1a and 1b: Histograms of the frequency of SRT Total Recall and FN-N Total Free Recall as well as SRT 30-Minute Delayed Recall and FN-N 30-Minute Delayed Recall shows that the FN-N distribution of scores has less of a ceiling effect than the SRT scores. (Scales on the x-axis represent the highest score possible on the test)

Figure 1

Figure 1a and 1b: Histograms of the frequency of SRT Total Recall and FN-N Total Free Recall as well as SRT 30-Minute Delayed Recall and FN-N 30-Minute Delayed Recall shows that the FN-N distribution of scores has less of a ceiling effect than the SRT scores. (Scales on the x-axis represent the highest score possible on the test)

Figure 2

Scatterplot of age and global Aβ deposition. Line at 1.15 indicates the separation between amyloid negative and amyloid positive individuals based on a combined sample of 140 subjects representing the lower limit for PiB retention in AD patients.

Figure 3

Figure 3a to 3c: Performances on the FN-N Composite (3a) the SRT Memory Composite (3b) and the FN-O Composite (3c) were inversely related to PiB retention in frontal and PPCLP cortices but only the FN-N Composite reached statistical significance. Analyses were co-varied for age and AMNART IQ.

Figure 3

Figure 3a to 3c: Performances on the FN-N Composite (3a) the SRT Memory Composite (3b) and the FN-O Composite (3c) were inversely related to PiB retention in frontal and PPCLP cortices but only the FN-N Composite reached statistical significance. Analyses were co-varied for age and AMNART IQ.

Figure 3

Figure 3a to 3c: Performances on the FN-N Composite (3a) the SRT Memory Composite (3b) and the FN-O Composite (3c) were inversely related to PiB retention in frontal and PPCLP cortices but only the FN-N Composite reached statistical significance. Analyses were co-varied for age and AMNART IQ.

Figure 4

Logistic regression revealed a significant relationship between PiB positive classification and lower performance on the FN-N Composite.