For individuals with mobility impairments, using a manual wheelchair (MWC) is an effective and accessible form of both mobility and aerobic exercise [1, 2]. However, largely due to the highly repetitive and strenuous nature of MWC propulsion, a significant proportion of long-term users develop pain and injury in the upper extremities. As many as 73% of users experience shoulder pain and up to 55% suffer from carpal tunnel syndrome [3–7]. Prevention of such injuries may be achieved through learning safe and efficient propulsion technique. In new MWC users, propulsion is a novel motor skill that must be acquired through motor learning principles as this ideal technique is far from intuitive.
Ergonomics and propulsion biomechanics research have identified several indices of propulsion quality and many bear complex relationships with risk for upper limb pain and injury. For example, push frequency, defined as the number of pushes per second, has been positively correlated with shoulder pain as well as impaired median nerve function [8, 9]. Sawatzky et al. (2015) recommend that during steady-state MWC propulsion, this measure should be maintained as one push / second or less [10]. Contact angle, defined as the angle over which the hand is in contact with the MWC pushrim, should be maintained between 85–100º [10]. A large contact angle promotes lower push frequency, smoother pushes, and lower rate of rise of force [11–13]. However, absolutely maximizing contact angle may also engender dangerous consequences as this goal promotes movement at the extremes of upper limb ranges of motion and repeatedly places joints at awkward and potentially injurious positions [14, 15].
Additionally, the clinical practice guidelines from the Consortium for Spinal Cord Medicine recommend that MWC users use long, smooth pushes which limit the amount of peak force applied to the pushrim. The guidelines also recommend the adoption of a semi-circular pattern, in which the hand follows the path of the pushrim, drifting below the rim during the recovery phase of the push [15]. The semi-circular pattern is specifically preferred because it has been associated with greater push time to recovery time ratio, lower push frequency, low joint acceleration, and minimal abrupt changes in hand direction [16–18]. In summary, the ideal technique appears to involve a semi-circular propulsion pattern with low push frequency, relatively large contact angle, and limitations on force. However, this pattern does not appear to be naturally intuitive for most new MWC users [11, 19, 20].
Evidently, safe and efficient MWC propulsion technique is a complex, novel motor skill that must be acquired using motor learning principles. However, new users are afforded very little time for specific propulsion training with a clinician during rehabilitation [21]. Virtual reality (VR) systems and simulators may serve as a potential solution to this problem, while also offering unique advantages such as safety, motivation, and increased practice time with fewer clinicians required [22, 23]. The current evidence also suggests that training protocols using VR-based simulators are valid and effective for facilitating positive changes in manual, as well as power wheelchair performance [24–26].
VR-based training also presents as a unique opportunity to measure and provide feedback to the user about their performance in real time. Specifically, augmented feedback (AF) is defined as information that is provided in addition to intrinsic sensory information that is naturally perceived during practice [27, 28]. A number of studies support the contention that AF about propulsion biomechanics can effectively and reliably produce desirable change in contact angle and push frequency [14, 29–31]. However, the effect of AF on improvement of kinetic parameters such as peak force, fraction of effective force, and power output is less consistent [32–34]. Notably, all of these studies have been performed on wheelchair ergometers or treadmills without VR, and very few utilized delayed retention tests. Further, very little testing has been done to confirm effective transfer of acquired skill to over-ground propulsion.
During VR training, AF may also be manipulated in modality, content, and timing, among other factors in order to facilitate error detection and learning. For example, feedback may be provided concurrently or terminally (i.e., during or after a trial), as well as visually, audibly, or haptically. Other elements of AF that may be varied are its relative frequency and delivery schedule. In contrast to predictions based on traditional motor learning theories, the guidance hypothesis asserts that feedback is useful early in practice but detrimental when relied upon for long periods of time [35]. This hypothesis predicts that a faded feedback schedule may be an ideal approach to learning as it captures the benefits of AF by guiding learners to correct movement patterns early in practice while avoiding dependence [36, 37]. In a faded schedule, AF is provided at high frequency early in practice and is gradually diminished as practice progresses. The guidance hypothesis has been robustly verified for training discrete, laboratory tasks such as manipulating a lever or producing target force waveforms [38–40]. However, it has not been confidently verified in more complex, ecological, and continuous tasks such as MWC propulsion.
When considering acquisition and learning of any motor task, it is imperative to emphasize the critical distinction between motor performance and motor learning. Motor learning refers specifically to long-term, relatively permanent changes in movement execution or capability. Meanwhile, transient improvements in motor performance may occur as a result of specific practice conditions [27]. For this reason, it is critical to administer delayed retention tests after sufficient time without practice in order to identify and measure true motor learning [41]. Furthermore, when training takes place in a controlled setting such as with a laboratory ergometer or in a VR simulator, it is also critical to consider whether acquired skills and techniques have real-world benefits. Wheelchair propulsion quality must also be measured over-ground before and after VR training in order to confirm that learning is not exclusively limited to simulator propulsion.
The purpose of the current study was to examine the effects of providing AF as well as its delivery schedule on motor learning, retention, and transfer of MWC propulsion technique as a novel motor skill. Specifically, this study sought to (1) compare the relative benefits of providing AF during training to conditions without AF, (2) compare a faded schedule to a high frequency schedule of AF, and (3) determine the transferability of technique acquired in a VR simulator.