Methods: We studied 45 normal subjects (mean age = 71.7±8.7, education = 16.6±2.7; AMNART IQ = 123.5±7.1) with clinical dementia rating (CDR) scores = 0 and MMSE≥28. Using multiple linear regression analysis, we related PiB retention (DVR, cerebellar reference) as a continuous variable in frontal, lateral temporal, lateral and medial parietal cortices, co-varying for age and AMNART IQ, to NP test scores. We transformed NP scores into composites representing traditional memory measures (all subtests of the Selective Reminding Test) and executive functions [Trails B, FAS, digits backward) as well as name retrieval on the FNAME.

Results: We found a significant inverse relationship for FNAME name retrieval with amyloid deposition in frontal (R2=0.28, β=-2.1, p=0.02), lateral temporal cortices (R2=0.270, β=-2.5, p=0.03), lateral and medial parietal cortices (R2=0.27, β=-2.4, p=0.05), that was not modified by IQ. Traditional memory measures did not show a significant relationship with amyloid deposition in these regions, except for frontal (R2=0.20, β=-0.30 p=0.04) and medial temporal regions (R2=0.26, β=-0.57 p=0.02) and only when IQ was included in the model. We did not observe any significant inverse relationship between frontal amyloid deposition and tests of executive function.

Conclusions: Performance on a demanding test of face/name retrieval, a common complaint of many normal older individuals, was associated with amyloid burden in brain regions critical for both executive function and memory systems, including frontal, temporal and parietal cortices. Challenging tests of associative memory retrieval in clinically, normal older individuals may prove to be sensitive markers for detecting early effects of amyloid deposition, regardless of cognitive reserve.