A single bout of high-intensity aerobic exercise facilitates response to paired associative stimulation and promotes sequence-specific implicit motor learning

Abstract

The objectives of the present study were to evaluate the impact of a single bout of high-intensity aerobic exercise on 1) long-term potentiation (LTP)-like neuroplasticity via response to paired associative stimulation (PAS) and 2) the temporal and spatial components of sequence-specific implicit motor learning. Additionally, relationships between exercise-induced increases in systemic brain-derived neurotrophic factor (BDNF) and response to PAS and motor learning were evaluated. Sixteen young healthy participants completed six experimental sessions, including the following: 1) rest followed by PAS; 2) aerobic exercise followed by PAS; 3) rest followed by practice of a continuous tracking (CT) task and 4) a no-exercise 24-h retention test; and 5) aerobic exercise followed by CT task practice and 6) a no-exercise 24-h retention test. The CT task included an embedded repeated sequence allowing for evaluation of sequence-specific implicit learning. Slope of motor-evoked potential recruitment curves generated with transcranial magnetic stimulation showed larger increases when PAS was preceded by aerobic exercise (59.8% increase) compared with rest (14.2% increase, P = 0.02). Time lag of CT task performance on the repeated sequence improved under the aerobic exercise condition from early (-100.8 ms) to late practice (-75.2 ms, P < 0.001) and was maintained at retention (-79.2 ms, P = 0.004) but did not change under the rest condition (P > 0.16). Systemic BDNF increased on average by 3.4-fold following aerobic exercise (P = 0.003), but the changes did not relate to neurophysiological or behavioral measures (P > 0.42). These results indicate that a single bout of high-intensity aerobic exercise can prime LTP-like neuroplasticity and promote sequence-specific implicit motor learning.

Keywords: aerobic exercise; brain-derived neurotrophic factor; motor learning; neuroplasticity.

Fig. 1.

Overview of experimental procedures. PAS, paired associative stimulation; CT, continuous tracking.

Fig. 2.

Schematic of the CT task. A: example of target movements during the random and repeated sequences within a single trial of tracking. Each black line depicts a different possible tracking pattern presented over 1 trial. The middle portion of the trial depicts the repeated sequence. 5 possible trials are depicted. B: participant view of target (○) and cursor (●) presented on computer monitor. C: participant manipulating joystick device with thumb of nondominant hand.

Fig. 3.

Changes in corticospinal excitability evoked by PAS preceded by rest and exercise in a single subject. A and B: motor-evoked potential (MEP) recruitment curves collected pre-PAS (gray) and post-PAS (black) for each condition (rest and exercise). C: average raw MEP waveforms (n = 10 MEPs) elicited pre-PAS (gray) and post-PAS (black) for each condition (rest and exercise) at each transcranial magnetic stimulation (TMS) intensity. RMT, Resting motor threshold.

Fig. 4.

Changes in corticospinal excitability evoked by PAS preceded by rest and exercise across the group. A: percent change in linear slope of MEP recruitment curves from baseline to pre-PAS and pre-PAS to post-PAS for the rest and exercise conditions when averaged across the group. Error bars represent 1 standard deviation. Horizontal bars and asterisks indicate statistically significant differences (P < 0.05). B: percent change in MEP recruitment curve slope from pre-PAS to post-PAS under the rest and exercise conditions for each participant. Unlinked data points represent the mean percent change across the group under each condition (as shown also by bar graph in A). Error bars around the unlinked data points demonstrate the 95% confidence interval for percent change in recruitment curve slope. C and D: MEP recruitment curves with MEP amplitudes averaged across the group for all TMS intensities pre-PAS and post-PAS for the rest and exercise conditions, respectively.

Fig. 5.

CT task data collected from a single subject when practice was preceded by rest (gray) and exercise (black). CT performance in terms of temporal error (lag behind the target) (A and B) and in terms of spatial error corrected for the contribution of time lag (shifted root mean squared error, RMSE) (C and D). For A and B, more negative lag values reflect greater temporal error. A and C demonstrate repeated sequence performance, whereas B and D show the random sequence performance. Solid lines demonstrate performance on each trial, and data points represent performance averaged over 3–4 trials. Vertical dashed lines indicate changes in the CT block (early practice, late practice, and 24-h retention block).

Fig. 6.

CT task performance in terms of temporal error (lag, A) and spatial error corrected for time lag (shifted RMSE, B) when averaged across the group. In A, more negative lag values reflect greater temporal error. B, right: effects of time and sequence on shifted RMSE detected by the statistical analyses. Early practice refers to the beginning of practice block 1, late practice to the end of practice block 2, and retention to the beginning of the 24-h retention block. Error bars represent 1 standard deviation. †Statistically significant difference from early practice (P < 0.05). Horizontal bars and asterisks indicate statistically significant differences between values (P < 0.05).