Who, what, where and how much: tDCS and training effects on working memory

Abstract

Working memory (WM) is strictly capacity limited and subject to age-related decline. Because WM is a fundamental executive function essential for most cognitive tasks many people are interested in expanding and/or stabilizing WM function. Whereas WM training alone provides mixed results, incorporating transcranial direct current stimulation (tDCS) offers promise across young and older adults. In several within-subjects designs we found that adults benefit from anodal tDCS applied to left or right PFC compared to sham tDCS. However, in single-session per montage studies only subsets of participants benefit. Notably, better educated older adults and higher WM capacity young adults benefit from left or right frontal tDCS, whereas lower WM capacity older adults benefit from lateralized right frontoparietal tDCS. Importantly from an applied perspective, longitudinal designs reveal consistent benefits across participants. Our data indicate that tDCS appears to prolong the return to baseline performance after WM training ends. We see near and far transfer in these multi-session designs suggesting effects have lasting effects on the neural networks underlying WM. There are optimization challenges associated with the large parameter space associated with tDCS. For instance, our data show that tDCS intensity can elicit non-linear benefits to WM indicating that more intensity is not always better. An added layer of complexity is that tDCS intensity interacts with a common single nucleotide point mutation in the catechol-o-methlytransferase (COMT) gene that metabolizes prefrontal dopamine and modulates WM capacity. To complicate matters, these interactions depend on WM task demands: visual or spatial. To optimize longitudinal tDCS for broad usage we must understand by what mechanism individual differences and genetics contribute to the lasting behavioral effects. In summary, once the contributions of individual differences (who), task (what), stimulation site (where) and intensity (how much) are identified, the translational value to a particular individual might be predictable.