Spaced Versus Massed Skill Learning

In cognitive psychology, practice is most effective when training sessions are distributed over time (spaced), rather than when they are close to each other (massed). This phenomenon, thought to engage long-term potentiation-like mechanisms in animal models and described as spacing effect, has not been investigated in the motor domain. It is not known if spaced motor training elicits longer lasting learning effects than massed motor training.

Objective and Study Population:

The purpose of this investigation is to assess the relevance of the spacing effect in motor skill learning in healthy volunteers and in patients with chronic stroke.

Design:

Experiment 1: Determination of long term learning in healthy volunteers with spaced and massed practice:

The first hypothesis is that spaced practice will enhance long-lasting learning of a motor task (defined as performance improvements measured 1 and 3 months post training) to a larger extent than massed practice in separate groups of healthy volunteers. Healthy volunteers will practice a well-characterized pinch force task following spaced or massed schedules in a factorial design (n=26). If this hypothesis is proven correct, we will proceed as suggested by PIRC, to Experiments 2 and 3, to gain information on the mechanisms underlying the superior training strategy in healthy volunteers (Exp 2) and to determine if this training strategy is also superior to massed practice in stroke patients (Exp 3), an issue of crucial importance in neurorehabilitation.

Experiment 2: Study of mechanisms underlying superior effects of spaced over massed practice in healthy volunteers, rTMS:

Previous work demonstrated the involvement of the primary motor cortex (M1) in consolidation of motor learning and the importance of top-down attentional control by the prefrontal cortex. It is possible that an enhanced recruitment of these two regions mediates the superior performance levels reached with spaced training. Here, we plan to study the effects of inhibitory 1 Hz TMS applied to M1 and PFC on performance improvements with spaced training. We hypothesize that the superiority of spaced practice, relative to massed practice, will be cancelled by down regulation of activity in M1 and PFC but not by sham or posterior parietal cortex (PPC) stimulation (n=104).

Experiment 3: Determination of long term learning in stroke patients with spaced and massed practice:

We hypothesize that motor learning in chronic stroke patients will improve to a larger extent with spaced than with massed practice (n=42).

This study is expected to delineate the role of the spacing effect on human motor learning and identify two of the possible neural cortical substrates in healthy volunteers and its possible beneficial effects on motor learning after stroke.

Experiment 4: Determination of whether original findings with the spacing effect in explicit motor learning generalized to implicit/procedural motor learning:

The ability to generalize to implicit motor learning is important because implicit learning (also known as procedural learning), which is defined as learning which occurs without awareness and without intention, underlies the development of automaticity which characterizes all well-learned motor skills (Reber, 1993; Squire, 2004). Hence, for the purposes of stroke rehabilitation, it is important to determine if and how the spacing effect occurs for implicit (procedural) learning which underlies the development of automaticity that characterizes all well-learned motor skills. In addition, using explicit motor sequencing tasks, it is difficult to determine if the spacing effect helps general motor skill or sequence-specific skill. In other words, performance benefits from the spacing effect can derive from improvements in the visuomotor transformation required to push the keys on the board independently of the presence or absence of sequence, or from motor sequencing improvements. Both of these issues can be addressed using the serial reaction time task (or SRTT), a well-studied implicit motor sequence learning task which is described in more detail in this protocol (Nissen and Bullemer, 1984).

We hypothesize that the spacing effect occurs in a sequence-specific manner for implicit motor sequence learning. We hypothesize that an SMA-based motor network underlies the superiority of the spacing effect for implicit motor sequence learning. We will show this by using 1Hz TMS to create virtual lesions and establish a cause-effect link between the SMA (or M1 but not CZ or sham) and superior motor skill with Spaced over Massed training (n=80).

Outcome Measures:

The primary outcome measure will be improvement in pinch force. Secondary outcome measurements will be measures of motor cortical excitability including motor evoked potential amplitudes, intracortical inhibition and facilitation.

For procedural/implicit motor sequence learning, the primary outcome measure will be an improvement in skill as seen by a difference in reaction time between sequenced and randomly ordered trials.