Transient Enhancement of Fluid Intelligence induced by Cortico-Cortical Paired-Associative Magnetic Stimulation

Fluid intelligence (Gf) represents the ability to cope with novel scenarios irrespective of previously acquired knowledge. Given its impact in everyday living, its correlation with academic and professional success, as well as... [ view full abstract ]

Fluid intelligence (Gf) represents the ability to cope with novel scenarios irrespective of previously acquired knowledge. Given its impact in everyday living, its correlation with academic and professional success, as well as its resilience to pathological conditions, improving Gf has become a crucial topic for contemporary neuroscience. Here we applied a novel Non-Invasive Brain Stimulation (NIBS) technique based on Hebbian plasticity principles by means of double-coil Transcranial Magnetic Stimulation (TMS), called cortico-cortical paired associative stimulation (cc-PAS). Concurrent stimulation of two Gf-related brain regions was delivered to induce a plastic rearrangement of inter-regional connectivity, possibly leading to a transient modification of individual Gf levels. 

Fifty healthy individuals performed an ad-hoc abstract reasoning task (i.e. modified SANDIA task) allowing to test both relational and logical reasoning abilities, and were randomly assigned to five stimulation groups. cc-PAS group carried out the task before and after a fully neuronavigated cc-PAS protocol. Stimulation was applied to a left-lateralized fronto-parietal network based on task-fMRI meta-analysis maps and individual functional connectivity (FC) fMRI data. Most specifically, the left inferior parietal sulcus (IPS) and left middle frontal gyrus (MFG) were targeted, being the most positively correlated nodes in the Gf FC network. Each session included 180 pairs of stimuli delivered at a rate of 0.2 Hz over a period of 15 min (Inter Stimulus Intervals (ISI)= +10ms, -10ms, 0ms). TMS intensity was based on resting motor threshold (RMT) as defined for each participant for the left primary motor cortex. To control for the local effects of TMS, participants also received either real and placebo (Sham) single-coil TMS over the MFG and IPS. Finally, in order to control for learning effects induced by repeated assessment of Gf, no-TMS control group of participants performed SANDIA four times in a row while not receiving any stimulation.

Depending on ISI, an increase in accuracy at the Sandia matrices was observed after cc-PAS. Specifically, we observe an increase in accuracy of approximately 5% for Relational items after STDP-inducing cc-PAS (i.e. cc-PAS +10ms, minus10ms), while a lack of gain in performance was observed while delivering no-delay (i.e. cc-PAS at 0ms) and single-coil TMS . No significant changes in reaction times or speed-accuracy trade-off were documented for Logical items.

The present results constitute an original evidence of cc-PAS effects on higher-order cognition in healthy individuals. Even though the complete understanding of the neurophysiological substrates of cc-PAS requires more in-depth electrophysiological investigations, the specificity of the present results suggests the possibility of modulating specific inter-regional brain dynamics in a timing-dependent manner. Data also strengthen the pivotal role of the left fronto-parietal network in Gf-related processing, promoting cc-PAS as a tool to causally validate the importance of specific cortical networks into explaining variability in Gf.