Cognitive training and retest learning effects on theta and alpha power in older and young adults: A perspective on the crunch hypothesis and the STAC-R model

Abstract

According to the STAC-R model, scaffolding enhancement is achievable through various interventions. Indicating forms of compensatory scaffolding, the STAC-R model refers to phenomena described in other theoretical models, such as the enhanced fronto-parietal recruitment described in the CRUNCH hypothesis. The presented study investigated whether working memory training can induce compensatory scaffolding in older adults through increased prefrontal and parietal involvement (indicated by changes in theta and alpha power). The sample comprised 90 individuals, including 45 participants from the experimental (22 older and 23 young adults) and 45 from the passive control group (21 older and 24 young adults). The age range was 60–75 years for older adults and 20–35 years for young adults. We assessed the effects of a 12-session working memory training with the use of the adaptive n-back task on theta and alpha power measured in frontal midline and central-parietal areas by EEG in older and young adults during the n-back task performance at three difficulty levels. At the behavioral level, we found a positive, significant improvement in cognitive performance in young adults from experimental group. In contrast, the positive changes in older adults were too small to prove statistically significant. At the level of neuronal activity, we observed not a training effect but a retest effect. It was revealed primarily for theta oscillations in older adults and manifested by increased theta power with higher task demands and equalization of theta power of older and younger persons in the post-test. For alpha oscillations, the retest effect was negligible, and its only manifestation observed in older adults was a reduction in the dependence of alpha power on task difficulty. The study results indicate limited potential for improving WM performance in older adults compared to young adults. The presence of the retest learning effect, instead of the training effect, proved that familiarity with the task was crucial, rather than regular training of its performance. Changes observed in older adults in theta power can be considered positive, and these results are consistent with the CRUNCH hypothesis of a compensatory role for increased executive control involvement. In turn, changes in the alpha power in the same group should be considered rather maladaptive. Nevertheless, given the overall study findings, it can be concluded that although the behavioral effects of training are stronger in young adults, the changes in neuronal activity resulting from the retest learning effect are more marked in older adults.