Double dissociation of working memory load effects induced by bilateral parietal modulation

Transcranial magnetic stimulation and neuroimaging data have revealed bilateral posterior parietal cortex (PPC) involvement during verbal n-back working memory (WM). In this task as n (i.e., WM load) increases, subjects show poorer behavioral performance as well as greater activation of this brain area. Moreover, there is evidence that a brief period of practice or even increased familiarity with the task can improve WM performance and lead to activation changes in the PPC. The aim of this study was to investigate, using transcranial direct current stimulation (tDCS), the effects on WM load performance induced by different PPC modulation after increased familiarity with the task. After a short practice, we tested verbal WM using an n-back task (1-back vs. 2-back) before and after the application of bilateral tDCS over PPCs (left anodal-right cathodal, left cathodal-right anodal or sham). ANOVA showed a significant interaction between tDCS and task. In the 1-back task, left anodal-right cathodal modulation abolished improvement in reaction times observed in the other two modulation conditions. Conversely, in the 2-back task the same effect was observed after left cathodal-right anodal modulation relative to the other two modulation conditions. This double dissociation demonstrates either a differential engagement of each PPC or changes in the interhemispheric balance of activity across this brain region. Neuroimaging studies show parametric activation of the PPC as difficulty increases, but activation does not switch sides. Thus, our observed effects cannot be attributed to increased task difficulty, the stimuli used, or the response requirements. Rather, we suggest that these findings reflect the use of different processing strategies to perform these two tasks. In conclusion, after increased familiarity with the task, different tDCS modulations lead to changes in a task-related region depending on differences in processing strategies in 1-back vs. 2-back.