The presence of after-effects in adaptation tasks implies that an existing internal model has been updated. Previously, we showed that although observers adapted to a visuomotor perturbation, they did not show after-effects. In this experiment, we tested 2 further observer groups and an actor group. Observers were now actively engaged in watching (encouraged through imagery and movement estimation), with one group physically practising for 25% of the trials (mixed). Participants estimated the hand movements that produced various cursor trajectories and/or their own hand movement from a preceding trial. These trials also allowed us to assess the development of explicit knowledge as a function of the three practice conditions. The pure observation group did not show after-effects, whereas the actor and mixed groups did. The pure observation group improved their ability to estimate hand movement of the video model. Although the actor and mixed groups improved in actual reaching accuracy, they did not improve in explicit estimation. The mixed group was more accurate in reaching during adaptation and showed larger after-effects than the actors. We suggest that observation encourages an explicit mode of learning, enabling performance benefits without corresponding changes to an internal model of the mapping between output and sensory input. However, some physical practice interspersed with observation can change the manner with which learning is achieved, encouraging implicit learning and the updating of an existing internal model.

During motor adaptation learning, consecutive physical practice of two different tasks compromises the retention of the first. However, there is evidence that observational practice, while still effectively aiding acquisition, will not lead to interference and hence prove to be a better practice method. Observers and Actors practised in a clockwise (Task A) followed by a counterclockwise (Task B) visually rotated environment, and retention was immediately assessed. An Observe-all and Act-all group were compared to two groups who both physically practised Task A, but then only observed (ObsB) or did not see or practice Task B (NoB). The two observer groups and the NoB control group better retained Task A than Actors, although importantly only the observer groups learnt Task B. RT data and explicit awareness of the rotation suggested that the observers had acquired their respective tasks in a more strategic manner than Actor and Control groups. We conclude that observational practice benefits learning of multiple tasks more than physical practice due to the lack of updating of implicit, internal models for aiming in the former.

Observers can learn to move in novel, adapted environments after watching a learning or expert model. Although this is an effective practice technique, it is unclear how this learning is achieved and if observers update an internal model of their visual-motor environment, as shown through the presence of after-effects (i.e., negative carry-over effects when aiming in a normal environment following exposure to perturbed conditions). For such updating to occur via observational practice, it has been reasoned that the observer requires the motor capabilities to perform the task they are observing. To test this, we first trained three groups to physically move in clockwise (CW) or counterclockwise (CCW) rotated environments. When immediately returned to a normal environment, after-effects were seen. We then attempted to wash out these effects before allowing two of these groups (CW and CCW), and a naïve observation only group, to watch a video of an actor performing in a CW environment. This observation phase was immediately followed by another test for after-effects and a direct test of learning when aiming in the rotated environment. Consistent with previous data, there were direct learning effects due to observation. Although after-effects increased for the experienced observers, these were small and were not significantly different from a physical practice only group that did not undergo the observation phase. Therefore, even with a motor repertoire for the rotated environment, there was a lack of evidence that observational practice results in implicit (re)updating of an internal model for aiming.

Visuomotor adaptation to novel environments can occur via non-physical means, such as observation. Observation does not appear to activate the same implicit learning processes as physical practice, rather it appears to be more strategic in nature. However, there is evidence that interspersing observational practice with physical practice can benefit performance and memory consolidation either through the combined benefits of separate processes or through a change in processes activated during observation trials. To test these ideas, we asked people to practice aiming to targets with visually rotated cursor feedback or engage in a combined practice schedule comprising physical practice and observation of projected videos showing successful aiming. Ninety-three participants were randomly assigned to one of five groups: massed physical practice (Act), distributed physical practice (Act+Rest), or one of 3 types of combined practice: alternating blocks (Obs_During), or all observation before (Obs_Pre) or after (Obs_Post) blocked physical practice. Participants received 100 practice trials (all or half were physical practice). All groups improved in adaptation trials and showed savings across the 24-h retention interval relative to initial practice. There was some forgetting for all groups, but the magnitudes were larger for physical practice groups. The Act and Obs_During groups were most accurate in retention and did not differ, suggesting that observation can serve as a replacement for physical practice if supplied intermittently and offers advantages above just resting. However, after-effects associated with combined practice were smaller than those for physical practice control groups, suggesting that beneficial learning effects as a result of observation were not due to activation of implicit learning processes. Reaction time, variable error, and post-test rotation drawings supported this conclusion that adaptation for observation groups was promoted by explicit/strategic processes.

Introduction: Stroke is a leading cause of long-term disability and can result in cognitive and motor impairments. Exercise may improve cognition and motor function after stroke, but past research has typically targeted these impairments in isolation. Here, we investigated whether pairing multiple bouts of exercise with motor practice can positively affect both cognitive and motor function after stroke. Methods: Thirty-three individuals with chronic stroke and 41 healthy older adults completed 5 separate days of motor task practice using their paretic/non-dominant arm, paired with 23 minutes of either high-intensity interval training exercise or rest. Cognitive and motor function were tested pre- and post-intervention. Processing speed, visuospatial skills, and inhibitory control was tested with the Trail Making Test-A and B (TMT-A, TMT-B) and object hit and avoid. Total completion time from TMT-A and TMT-B, target and distractor hits from object hit and avoid were used as dependent measures for assessing cognitive function. For individuals with stroke, motor impairment and function were tested with the Fugl-Meyer upper extremity arm assessment and Wolf Motor Function Test respectively. Results: All participants showed evidence of motor learning; exercise did not confer an additional benefit beyond that stimulated by practice for either group. For stroke participants, motor function (p = .047), but not motor impairment, improved over time. The stroke group who exercised before motor practice displayed significant reductions in TMT-A completion time (p = .035). Both stroke and older adults hit more targets in the post- compared to pre-intervention (p < .001), driven by improvements in the affected/non-dominant hand. Importantly, exercise paired with motor practice also led to a reduced number of distractors hit (p = .026) in the object hit and avoid task for both individuals with stroke and older adults. These changes were not at the expense of speed. Discussion: Five days of high-intensity interval training exercise paired with motor practice led to improved processing speed as measured by the TMT-A for individuals with stroke. Both exercise participant groups (stroke and older adults) showed improved visuospatial skills and inhibitory control as measured by object hit and avoid following our intervention. Our findings suggest that exercise paired with motor task practice leads to improved cognitive-motor function in individuals with stroke and older adults. Together, exercise paired with skilled motor practice appears to be a safe and effective means of enhancing cognitive-motor skills after stroke and in older adults.