Virtual reality (VR) has emerged as a promising tool for psychological and movement-based rehabilitation. For example, VR has been used to improve gait adaptability and stability in populations with mobility impairment and a heightened risk for falls [1,2], alleviate phantom limb pain in patients with upper extremity amputation [3], reduce combat-related posttraumatic stress in active duty service members [4], and improve proprioception, mobility, and muscle strength in older adults with knee osteoarthritis [5]. A clear advantage of the VR environment is that it can provide a gamified intervention that is designed to increase enjoyment, motivation, and retention. This can be particularly beneficial when the goal of the intervention is to stimulate movements that may be associated with pain and fear. Based on these unique advantages and the increase in user-friendly VR systems that continue to reduce in cost, it is likely that VR could become a fundamental component of psychological and movement-based rehabilitation programs.

Our group has been developing and testing novel VR games to assess and improve movement deficits in patients with low back pain (LBP) [6]. LBP is the most common source of pain among middle-aged and older adults, resulting in significant financial impact through both health care costs and pain-related work absences [7,8]. Much of these health care costs are driven by the approximately 10% of patients who develop chronic LBP that lasts for many years [9]. Many patients with LBP will develop kinesiophobia, which is fear of movement due to expectations of pain and harm [10]. Among patients with LBP, kinesiophobia is associated with reduced physical activity and particularly reduced spine motion and is a strong predictor of chronic LBP development [10-14]. In brief, kinesiophobia is central to the development of chronic pain through a vicious cycle of pain catastrophizing, hypervigilance regarding pain-related threat, fear-related avoidance of movement, and a resultant combination of disuse, physical deconditioning, and depression, which serve to further amplify pain [10,11,13].

A common method to quantify avoidance behavior is through the assessment of motor control during functional tasks that involve kinematic redundancy [12-14]. Our lab uses a standardized reaching task to compare forward reaching across participants with various body anthropometrics [12-14]. This standardized reaching task uses hip height, trunk length, and arm length to compute 4 points in space that would require exactly 15°, 30°, 45°, and 60° of trunk flexion (θ) to reach, if the participant were to move only at the trunk (Figure 1). While these points are computed based on isolated trunk flexion, participants are not instructed as to how they should reach each target. By assessing how an individual chooses to reach each target, this kinematically redundant task is well positioned to identify avoidance behavior (eg, reduced trunk motion and increased motion at the ankle, knee, and hips) [12-14]. While we originally implemented this reaching task with physical targets, we have recently developed a VR version of the task, hereby referred to as “Reachality” (Figure 2A) [15]. In Reachality, participants are immersed in a virtual environment using a head-mounted display, where their body is represented as an avatar and the participants are instructed to reach their hand through targets that appear in front of them (eg, at the 4 aforementioned locations).

Common interventional approaches for patients with LBP and a high fear of movement include graded exposure therapy, wherein patients gradually confront increasingly feared movements, and motor control exercises, wherein existing movement patterns are retrained with a specific focus on restoring trunk control [16-18]. Our group recently developed a VR dodgeball game, hereby referred to as “Dodgeality,” with the goal of creating a gamified intervention that incorporates principles of these classic interventional approaches that are known to be successful [6,19]. In Dodgeality, participants are immersed in a virtual gymnasium and hold a plastic 3D-printed dodgeball, which is tracked and visualized as a dodgeball in the avatar’s hand within the virtual environment. There are 4 opposing players who randomly take turns throwing dodgeballs at the participants, who are instructed to block the balls thrown at them with the ball held in their hand. Each dodgeball is launched at a constant velocity and is modeled as a point-mass projectile with only gravitational forces acting upon it, resulting in a parabolic flight path. Through manipulating only launch angle, we are able to force the dodgeballs’ trajectory to intercept the same 4 locations in space that are presented as targets during Reachality (Figure 1), thereby enticing the participants to reach those locations during gameplay. In a recent phase I clinical trial, our group found that lower intercept locations resulted in increased lumbar flexion excursion in patients with chronic LBP [6]. As avoidance behavior is more prominent when reaching for lower targets [14], the ability to increase the magnitude of trunk flexion needed for gameplay throughout an intervention is essential. Other important findings from the phase I clinical trial were that participants viewed the game as “distracting from their back pain” and “fun to play,” that they were “not worried about hurting their back during gameplay,” and that the game “did not increase their back pain” [6]. Based on the findings from this phase I trial, we believe that Dodgeality could be a useful component of rehabilitation for patients with chronic LBP and a high fear of movement, and we are currently evaluating the efficacy of this intervention in a phase II randomized clinical trial [20].

While our recent findings indicate that we are able to successfully manipulate the amount of trunk flexion needed during gameplay [6], Dodgeality is inherently a fast-paced game that requires quick reactions. In open-ended survey responses, some participants indicated a desire for more practice in the virtual environment at slower speeds [6]. Importantly, a primary principle of graded exposure therapy is to begin with patients’ least feared movement and gradually work toward their most feared movement [16-18]. Patients with LBP and a high fear of movement not only restrict the amount of trunk motion during forward reaching but also the speed at which they flex their trunk [12]. Therefore, to improve our ability to gradually increase mechanical demands of the trunk throughout an intervention, we sought to create a new game that required similar amounts of trunk flexion as Dodgeality but with lower velocity demands. We developed a virtual fish-catching game, hereby referred to as “Fishality,” which is intended to precede Dodgeality in a graded intervention (Figure 2B-2D). In the Fishality virtual environment, participants are standing on a dock overlooking a pond and have a basket in their dominant hand (which is tracked with a controller in the real world). Fish swim toward participants, with an indicator above the level of the water to alert participants to an approaching fish. The fish then jump out of the water toward participants at a high parabolic arc, and participants are instructed to catch the fish in their bucket. This trajectory is intended to give participants more time to react, thereby requiring decreased trunk flexion velocities. Through prescribing the height of the trajectory and resolving the kinematic equations, initial launch angle and launch velocity can be computed to ensure that the fish intercept the same 4 points in space that are used during Reachality and Dodgeality (Figure 1).

The purpose of this study was to compare lumbar kinematics across Reachality, Dodgeality, and Fishality in healthy control participants. Our first hypothesis was that lumbar flexion velocity would be increased during Dodgeality relative to that during Fishality. Our second hypothesis was that the extent of lumbar flexion would not be different between virtual games. While Dodgeality and Fishality are designed such that the trajectories of the launched objects intersect each of the 4 static target locations presented during Reachality, the participants are allowed to intercept the launched objects at any point along their trajectory. Considering that the objects’ trajectories are markedly different between Dodgeality and Fishality, it is possible that observed differences in lumbar kinematics may be explained through differences in the actual interception location (rather than the intended location from which trajectories are initially derived). Therefore, we present the following exploratory third hypothesis: participants would reach further in the forward direction during Fishality and Reachality than during Dodgeality.

The raw data and results for the repeated measures analysis of variance are presented in Table 1. There was an interaction between game and impact height for each outcome (Figure 3, all P<.001). Therefore, post-hoc simple effects models (ie, 1-way analysis of variance) were performed to compare the 3 games across each impact height.

Gamified movement-based intervention is a promising approach for rehabilitation in patients with LBP and a high fear of movement. Our group recently developed Dodgeality, a virtual dodgeball game where patients are encouraged to bend forward to block incoming balls thrown at them by opposing players [6]. While we are able to influence the amount of trunk flexion needed for successful gameplay by modifying the intended impact location of launched balls, Dodgeality is inherently a fast-paced game with large movement velocity demands. We sought to supplement our VR game suite by developing a game to encourage patients to flex their trunk at a slower speed. We therefore developed Fishality, a novel VR game where fish jump out of a body of water with a high parabolic arc, and the patients have to bend forward to catch the fish before they land back in the water.

This study sought to compare movement biomechanics across Dodgeality, Fishality, and a standardized virtual reaching task (Reachality). Our first hypothesis was supported, as lumbar flexion velocity was greater during Dodgeality than during Fishality. While flexion velocity during Fishality was less than that during Dodgeality at each intended impact height, differences were greater for higher impact heights (requiring less motion). Specifically, lumbar flexion velocity reduced by 38% at intended impact height 1 (Dodgeality: mean 71.5°, SD 35.8°; Fishality: mean 44.3°, SD 15.4°) and 21% at intended impact height 4 (Dodgeality: mean 88.7°, SD 35.3°; Fishality: mean 70.2°, SD 23.7°). Our second hypothesis was not supported, as lumbar motion was different between games. Specifically, Fishality resulted in 13%-18% less lumbar motion relative to Dodgeality for higher targets, resulting in magnitudes of movement that were more similar to those for Reachality. It is unclear why lumbar flexion was increased for higher targets during Dodgeality; however, it is likely that participants began moving downward before they identified the target of the incoming ball because of the fast speeds at which the dodgeballs were launched. This finding suggests that Fishality is better than Dodgeality for manipulating trunk flexion during gameplay. As the magnitude of lumbar flexion and lumbar flexion velocity across VR games and impact heights were comparable between this study and prior research conducted in a real-world environment [12,21], the findings of this study were likely not due to the VR environment itself. Additional evidence for this notion comes from a prior study that found limited differences in lumbar motion and velocity when reaching tasks were compared between a virtual environment and real-world setting [15]. Because a major goal of graded intervention is to increase movement demands throughout the course of an intervention [16-18], the ability to modulate lumbar motion and velocity is essential. Hence, the results of this study are encouraging when considering the development of an individualized, graded intervention program for patients with LBP who have low physical activity and a high fear of movement.

Another important finding from this study was that participants did not reach as far forward when playing Dodgeality and Fishality relative to when making contact with the targets during Reachality. This finding intuitively makes sense, as projectiles (dodgeball and fish) can be intercepted at any point along their trajectory for successful gameplay during Dodgeality and Fishality (compared to static target positions used in Reachality). Reach distance was increased during Fishality compared to that during Dodgeality, which is also to be expected given the different requirements of the 2 games. Specifically, incoming dodgeballs have a flat trajectory that will contact participants’ bodies if not blocked, whereas incoming fish have a high parabolic trajectory and will land in the water between the participant and intended intercept target if not caught. Based on these findings, feature modifications to our VR games, such as interception boundaries, could be introduced to ensure greater forward movement during gameplay, which would improve our ability to specifically target lumbar flexion appropriately across all target heights.

This study had limitations that should be considered when interpreting the findings. First, as our sample consisted of young healthy participants, these findings should be repeated in a cohort of participants with a broad range of ages and spine impairments to determine the robustness of the findings. However, for this study, we intentionally included healthy participants without impairment to ensure that the task demands aligned with how we had developed the VR games. A second potential limitation is that the gameplay order was not randomized, which could have introduced an ordering effect into our data. However, as the games are designed to be used in an ordered fashion for interventional purposes, we wished to investigate movement behaviors within this context.

In conclusion, the present study sought to compare a virtual dodgeball game, a newly developed virtual fish-catching game, and a virtual reaching task in a healthy sample. We found that lumbar flexion velocities were reduced in Fishality compared to those in Dodgeality and resembled velocity demands more similar to those in a self-paced reaching task (ie, Reachality). These findings are encouraging for future research aimed at developing individualized, graded VR interventions for patients with LBP and a high fear of movement.