Fig 1.
Motor learning task and experimental paradigm.
(A) Sequential Visual Isometric Pinch Task (SVIPT). Participants were instructed to move the cursor (black square) as quickly and accurately as possible between the “Home” position and each of the five gates by varying the amount of force applied to a force transducer. (B) Sample skill measure change during SVIPT training, shown as a shift in speed-accuracy tradeoff function (SAF). Note that as movement duration and/or error rate improve, SAF exhibits a downward shift resulting in increased skill measure. (C) Overview of study design. Participants in Aerobic and Aerobic + Rest groups ran at 65–85% of their age-predicated maximum heart rate for 30 minutes either immediately (Aerobic) or one hour (Aerobic + Rest) prior to motor skill training. Participants in the Control groups walked at a slow pace (1.0 m/s) for 35 minutes immediately prior to motor skill training.
Table 1.
Fig 2.
SVIPT performance for single day experiment.
Exercise intervention, shown as a grey bar, occurs between blocks 1 and 2 of SVIPT training. (A) Learning curve during motor training. Repeated-measures ANOVA revealed a significant main effect of Group (p = .021) and a Block x Group interaction (p = .007) for skill measure, indicating improved motor acquisition in the Aerobic group when exercise was performed immediately before motor skill training. (B) Error rate and (C) movement duration. Results indicate that the effect of exercise on skill measure may have been driven by a non-significant improvement in error rate (p = .127), as opposed to movement duration (p = .828). All data shown as mean ± SEM.
Fig 3.
SVIPT performance for multiple day experiment.
Experimental sessions took place on four subsequent days. Exercise intervention was performed immediately prior to SVIPT training on days 1–3, and only SVIPT training was performed on day 4. (A) Learning curve across all four days. Aerobic exercise prior to motor training led to an immediate and persistent improvement in skill measure compared with Controls (Group: p = .011) throughout exercise days. On the non-exercise day, Aerobic subjects tended to continue to exceed Controls in both overall performance (Group: p = .081) and learning rate (Block x Group: p = .065). No significant offline changes (i.e. retention) of skill measure were found for either exercise or non-exercise days. (B) Error rate and (C) movement duration. On exercise days, results indicate that the exercise intervention led to a significantly lower error rate in the Aerobic group (p = .006), but had no effect on movement duration (p = .479). A Group effect (p = .013) and Block x Group interaction (p = .022) on error rate were found on the non-exercise day as well. No significant offline changes in error rate or movement duration were found for either exercise or non-exercise days. All data shown as mean ± SEM.