Two previous studies demonstrated greater cognitive decline in MCI patients with positive amyloid-PET compared to negative MCI, as assessed by several neuropsychological tests, over 18 (Doraiswamy et al., 2012) and 36 months (Doraiswamy et al., 2014) follow-up. Our aim is to evaluate whether positive amyloid-PET scans are predictive of greater cognitive decline than negative scans over a shorter 12 months period.


      A naturalistic series of seventy-six MCI patients underwent amyloid-PET with 18F-Florbetapir (FBP-PET), and homogeneous clinical and neuropsychological multi-domain assessment before FBP-PET and after 12 months (12-FU). For each neuropsychological test we constructed a linear mixed model with scores as the dependent variable, and ‘FBP-PET result’ (positive or negative) and ‘time’ (baseline or 12-FU) as main factors. ‘Age’ and ‘education’ were included as covariates. Effect size indices (partial Eta-squared, ηp2) were also computed (0.02, 0.13 and 0.26 referring respectively to small, medium and large Eta-squared).


      Patients with positive scan had overall worse performance on tests of global functioning (MMSE, p=0.005, ηp2=0.11; ADAS-COG, p=0.007, ηp2=0.10) and on tests involving memory (Story recall, p=0.002, ηp2=0.13; Rey AVLT – Delayed recall, p=0.005, ηp2=0.11; Rey-Osterrieth complex figure – Recall, p = 0.001, ηp2=0.15). A significant worsening over 12 months was observed in ADAS-COG (p=0.003, ηp2=0.12), Rey-Osterrieth complex figure – Copy (p<0.001, ηp2=0.19), Raven’s CPM (p=0.039, ηp2=0.06), Letter (p=0.044, ηp2=0.06) and Category fluency (p=0.021, ηp2=0.08). Finally, positive patients showed a significantly greater cognitive decline over the time in the Story recall test (p=0.026, ηp2=0.07) and the Rey-Osterrieth complex figure – Copy (p=0.018, ηp2=0.08) (Figure).


      The cognitive performance of MCI patients positive to amyloid-PET is overall worse than in negative ones, but significantly greater deterioration over 12 months emerged in only a minority of tests. Longer time or more specific episodic memory tests may be useful for more sensitive monitoring of performance. Differences with previous studies may be due to different selection of tests or to our shorter follow-up. These data contribute information to define the proper time window for repeated testing for neuropsychology used as a biomarker for AD.
      Figure thumbnail fx1
      FigureCognitive decline in MCI patients over a 12-month follow-up. Horizontal parentheses indicate significant effect of the factor ‘time’, vertical parenthese indicate significant effect of the factor ‘FBP-PET result’ and ‘*’ indicates significant interaction of these factors.