Cognitive training effects in mild cognitive impairment with and without amyloid pathology.

Imagery-based cognitive training (CT) is a promising strategy for individuals with amnestic mild cognitive impairment (aMCI). However, there is a lack of studies investigating its differential impact in aMCI with and without amyloid pathology. Objective: To investigate the effects of CT based on mental imagery on episodic memory in individuals with amyloid-positive aMCI (Aβ+), amyloid-negative aMCI (Aβ-), and healthy controls (HC). Methods: Fifty-five participants were included: Aβ+aMCI (n=19), Aβ-aMCI (n=20), and HC (n=16), aged 51-89 years, with more than 6 years of education. Amyloid status was confirmed by 11C-PiB PET. All participants completed pre- and post-training neuropsychological assessments and attended six individual online CT sessions. Performance was assessed using the logical memory test (LM), Rey Auditory Verbal Learning Test (RAVLT), and generalization tasks (GT I-II). Results: Data were analyzed using the Shapiro-Wilk, Levene, Wilcoxon, ANOVA, Kruskal-Wallis, and Bonferroni tests. Significant improvements in immediate episodic memory (LM I, RAVLT A1-A5) were observed in the Aβ+ and Aβ-groups, particularly in the Aβ-aMCI group. Delayed recall (LM II, RAVLT A7) increased, although without statistical significance. All groups showed gains in generalization tasks, supporting the benefit of CT. Conclusion: Unimodal CT based on mental imagery promoted gains in immediate episodic memory across all groups. Although delayed recall gains were not statistically significant, the strategy was internalized and generalized. Similar responsiveness in Aβ+ and Aβ-groups suggest preserved compensatory mechanisms despite amyloid pathology. These findings suggest more robust effects at the level of encoding than consolidation, which may vary according to task characteristics and underlying neural integrity.