This three-arm randomized controlled trial assigned participants to 8 weeks of cognitive training, divided into 3 groups: Reading, Origami, Control. Inclusion criteria were age 65 years or older and Montreal Cognitive Assessment (MoCA) score>26. MoCA, Free and Cued Selective Reminding Test (FCSRT), and electroencephalogram (EEG) with spectral analysis while performing memory tasks as objective measures of synaptic function.

The primary outcomes of the study were analyzed using a 2x3 repeated measures ANOVA, with time (pre- and post-intervention) and randomization group. Tests of simple effects and interaction contrasts were utlized to highlight the locus of this effect. Experiment-wise type I error rate at α=.05 was controlled with Fisher's least significant difference (LSD), and effects investigated only where the omnibus tests were significant. Data distributions and model residuals were examined to assess the tenability of model assumptions.

During memory tasks, response times decreased within both Origami (p=.002) and Reading (p=.008) groups, but not the Control group (p=.941). There was a statistically significant decrease in matching response time from pre- to post-intervention that was larger in both the intervention groups (p=.037) compared to control.

There was a significant group by time interaction for theta power in the right parietal lobe (p = .043). Theta power increased in the right parietal lobe for group Reading (p = .009). Differences for theta power in the left frontal lobe were non-significant. For alpha power in the left parietal lobe, there was a significant increase in alpha power for Reading group (p=.007).

Further studies are needed to investigate the potentially beneficial role of cognitive training interventions.