The effect of metabolic loading on statistical learning

We investigated whether concurrent exercise would affect statistical learning (SL). During familiarization, participants were exposed to pictures that appeared sequentially, in a seemingly random fashion. In fact, the pictures were grouped into triplets. In the surprise test phase, participants identified triplets they had seen during familiarization. There were three groups: a group that performed familiarization seated on an exercise bike (CON), a group that performed familiarization while engaged in resistance free cycling (RF), and a group that performed familiarization while cycling at 60% of maximum effort (EX). The CON group correctly identified 72% of triplets in the test phase. The RF and EX groups correctly identified 61% and 55%, respectively. Only the CON group demonstrated performance that was significantly greater than chance. The RF group only just failed to demonstrate significant SL. Thus, concurrent exercise can suppress SL. Work is underway to determine the mechanism by which such suppression occurs.