Annually, 1 in 4 older adults are injured from falling [1], and the incidence rates [2] and resulting morbidities continue to rise [3]. Alongside rapidly advancing virtual reality (VR) technology, scientists and clinicians are working to predict and prevent falls using a range of nonimmersive and immersive techniques [4-6]. Yet, researchers are only beginning to understand the potential benefits of VR technologies and their capabilities to target the perceptual, cognitive, and motor processes related to fall risk [7]. We believe there is a disconnect between how VR is currently used to understand and prevent falls in experimental settings and its capacity to identify and target the processes that are involved when older adults fall in daily life.

We argue that scientists currently using VR technology to evaluate and modify fall risk often overlook stressors such as anxiety or the “fear of falling” that are associated with higher prevalence of falls [8,9]. As we age, we use more attentional capacity while walking and stepping becomes less “automatic” [10]; thus, the added cognitive demand of fear of falling could be detrimental to balance and gait for older people [11-14]. Our underlying argument is that the context of balance control is important, and scientists should aim to create paradigms that better replicate the challenges that older adults experience during everyday walking to understand and prevent falls. By not representing the affective context that exists in the real world in VR-based tasks and interventions, we fail to address the interactions between perceptual, cognitive, and motor processes that underlie effective or maladaptive balance in daily life [15-18]. Scientists now have a unique opportunity to develop VR-based tools that facilitate interactions with realistic mobility contexts that induce anxiety as in the real world (eg, crossing a busy street at night and walking in a crowded mall). By using VR to engage with realistic contexts, we could better identify older adults who are most at risk of falling and develop refined interventions to prevent falls.

Our overarching interests are to establish early diagnostics that identify who is at greatest risk of falling and develop specific interventions that reduce the risk of falls in everyday life. In this perspective paper, we discuss leveraging the rising popularity of VR to develop meaningful protocols for diagnosing and treating the risk of falling in older adults. Specifically, we discuss how VR simulations can recreate the typical perceptual, cognitive, motor, and affective demands of daily life to facilitate diagnostics and interventions that reduce the risk of falling. The key points discussed in this paper are shown in Textbox 1.

As a result of poor “balance confidence” or low “falls efficacy,” many older adults report a “fear of falling” and experience mobility-related anxiety during locomotion [16]. We provide a brief overview of how fear and anxiety influence fall risk (ie, the physiological and cognitive response to a perceived threat in a balance or walking task). We refer the reader elsewhere for detailed reviews that distinguish these processes from concepts such as concern about falling [13,16,18]. Older adults who experience a fall are more likely to exhibit gait impairments and suffer higher incidents of falling [19,20]. We define “mobility-related anxiety” as a phobia specific to walking that is associated with increased physiological arousal and cognitive stress, which interferes with perceptual, cognitive, and motor processes during walking in young and older adults [12,21-24]. Both self-reported fear of falling and mobility-related anxiety are believed to impair cognitive-motor control, predisposing older adults to a greater risk of slips, trips, and falls [13,14,18]. However, there may be some protective element to mobility-related anxiety, encouraging older adults to compensate for poor balance by “consciously monitoring” their actions in efforts to remain safe [17]. The degree that mobility-related anxiety impedes or protects older adult balance is still unclear because current evidence is only supported by cross-sectional associations between fear of falling and motor outcomes [25], fall incidences [26], or anxiety-related responses from experimental manipulations (eg, raising participants on a platform) [11,23,27]. The experimental tasks used by scientists and clinicians are strictly controlled, and consequently, these tasks may not induce anxiety in a way that represents the anxiety-inducing experiences of older adults in real-world mobility scenarios (eg, crossing a busy street at night). We argue that scientists should leverage advances in technology to develop representative mobility tasks within VR simulations while simultaneously retaining experimental control. VR presents an ideal tool to create realistic contexts that would enhance and individualize fall-risk detection and prevention.

Scientists interested in the effects of anxiety on mobility have predominantly used VR to create controlled experiments that answer fundamental science questions. Unfortunately, often the experimental conditions induced have limited relation to situations encountered in daily life. For example, our previous work has shown that immersive VR can be used to increase mobility-related anxiety in healthy people by simulating standing on a wooden plank and raising the walkway approximately 15 meters above ground level [24,28]. Although our approach allowed us to answer important questions about anxiety and motor control, we are hesitant to equate our results to behaviors in daily life. Historically, obtaining results that are representative of everyday tasks, or “representative task design,” has been a persistent limitation across many fields [29,30]. We do not yet know if the experimental outcomes emerging from controlled, laboratory-based studies are representative of behavior in everyday life. However, existing VR technology lends itself to probing affective responses by systematically imposing mobility-related threats to stability and testing balance control across external task constraints.

Scientists have typically used VR to induce mobility-related anxiety by replicating laboratory-based studies that physically raise participants to elevated heights [23,27,31]. The aim is to stimulate affective systems by increasing the perceived consequences of falling, also known as “postural threat” [23,31]. Simulating elevated heights in VR elicits similar changes to state anxiety (ie, subjective ratings) and motor behavior (ie, smaller postural sway area) as in the real world [32-34]. Various types of threatening contexts can elicit anxiety-related responses, ranging from simulated heights in a replicated laboratory environment [33,35,36] to riding an open elevator [37] and walking across a deep pit [38,39].

Another approach to study mobility-specific anxiety is to increase the difficulty of the locomotor task by walking on a narrow path. By increasing attentional demand and the energy cost of gait patterns [40,41], a narrower path imposes a demanding locomotor constraint that requires participants to continuously adapt and monitor their steps. A narrower step width also involves intrinsic risk appraisal by challenging stability; participants must evaluate their balance and devote added attentional resources to compensate for their anticipated missteps [42]. We interpret the added cognitive load of taking narrower steps as analogous to the “conscious monitoring” experienced by older adults who are anxious while walking [13,43].

VR-based tools can combine environmental and task-specific influences on mobility-related anxiety, emphasizing their potential to systematically examine the interactions between key perceptual, cognitive, and motor behaviors. Walking in VR on broad and narrow roads, and at low and high elevations, can reveal complex interactions between anxiety, cognition, and motor behavior in young and older adults [44]. For instance, combining path width and height manipulations in VR increases motor difficulty in addition to multiplying to consequences of a fall, forcing participants to prioritize performing the task safely versus quickly, revealing the interactions between mobility-related anxiety and motor performance [44]. Challenging balance in threatening VR environments (an increasingly narrowing path) reveals that balance confidence mediates the efficacy of older adults’ stepping responses [45,46]. Cumulatively, these results highlight the salience of mobility-related anxiety in locomotor behavior and its potential importance in improving balance and reducing fall risk.

However, while VR has led to advances in understanding mobility-related anxiety and balance, there remains a disconnect between experimental manipulations of anxiety responses and the lived experience of older adults suffering from mobility-related anxiety in daily life. We propose that researchers use VR to simultaneously test the influence of task- and environmentally driven affective responses on locomotion, while working to develop generalizable applications. We envision VR could be used as a tool to generate outcomes that are experimentally controlled, while simultaneously being representative of everyday contexts that precipitate real-world falls.

Rehabilitation scientists are limited in their ability to replicate complex everyday scenarios and elicit representative anxiety, hindering the ecological validity and translation of existing approaches. Simultaneously, we have yet to find a way to measure perceptual, cognitive, and motor processes in the real world without using disruptive and resource-heavy equipment. To address both limitations, VR can simulate everyday situational demands and quantify responses with high resolution, thereby helping to bridge the gap between laboratory-based research outcomes and everyday fall risk. As opposed to evaluating physical function in sterile clinical settings with clear hallways, using VR to challenge older adult mobility within an ecologically valid setting would better reveal real-world mobility deficits, thereby potentially leading to enhanced therapies to reduce fall risk.

Crossing the street at night is an example of a locomotor context that elicits a specific mobility-related anxiety and requires effective perceptual (eg, judging the distance to oncoming hazards with low-quality visual feedback), cognitive (eg, gathering and retaining limited visual and spatial information in working memory to monitor position), and motor adaptations (ie, neuromuscular coordination and step targeting) to achieve safely. It may therefore be no surprise that older adults experience a higher prevalence of traffic-related injuries compared to younger pedestrians; a consequence attributed to a host of individual, task, and environmental factors [95], including self-reported fear of falling [96]. The capability of VR-based tools to replicate a street-crossing simulation for older adults was demonstrated in a series of studies using a projection-based room with a 360° simulation of a busy street and crosswalk [97], allowing people to move through a realistic 3D visual surround. The results show when older pedestrians avoid moving traffic, their cognitive-motor “workload” is higher than when crossing an empty street, leading older adults to walk faster and ignore traffic-avoidance–related tasks [98]. These latter behaviors are reminiscent of older adults in artificial anxiety-inducing settings (ie, a narrow or elevated walkway) [44]. A similar study recently used a street-crossing paradigm in an immersive commercial HMD apparatus measuring eye tracking, locomotor behavior, and cognitive performance to draw inferences about pedestrian behaviors in young adults [99]. Such studies highlight the potential to use VR to simulate realistic everyday contexts that challenge affective responses while being feasible for older adult users. We believe further efforts are needed to develop similar immersive, yet effective, VR-based approaches using simulations of real-world contexts where falls are common.

The greater challenge is to capture behavior from fearful older adults during threatening walking scenarios and best use VR to quantify behaviors that perpetuate falls. We propose a focus on scenarios where older individuals experience a fear of falling, especially in settings leading to real-world falls such as stairwells, escalators, crowded sidewalks, shopping malls, or nighttime streets. Borrowing from the concepts developed from successful trauma-focused VR-based treatment of posttraumatic stress disorder (PTSD) in soldiers [99,100], combat scenarios are predictable sources of phobic anxiety. Therefore, presenting a realistic combat scenario in VR is likely to elicit a phobic response. Yet, we cannot be certain which scenarios cause anxiety responses for older adults in real life and how those vary individually. We propose qualitative interview-based inquiry [101], a type of “needs assessment” or “task analysis,” as a step toward determining the scenarios that are anxiety inducing for older adults with motor impairments related to a concern about falling. Our previous qualitative work interviewed fearful older adults to examine their thoughts during hypothetical scenarios leading to falls, and our findings debunked popular theories about attentional focus (ie, attentional control theory) under anxiety or stress developed using data from athletes [102]. Whereas athletes are distracted by a "threat bias" in attentional control that is maladaptive in competitive settings [103], we showed that some degree of "conscious monitoring" dring walking is protective for older adults who realistically appraise their sense of control in the situation [17]. Only individuals who expressed lacking a sense of control demonstrated a maladaptive fear of falling [17], highlighting the value of qualitative inquiry in translating theory into population-specific interventions.

VR environments that represent everyday situations associated with a fear of falling would allow clinicians to examine locomotor performance on a range of functional tasks in a fall risk "stress test". Such a virtual “stress test” was recently demonstrated in a functional VR game using an urban scenario designed for rehabilitating activities of daily living [104]. The “UrbanRehab” VR tool was created to optimize ecological validity and challenge outdoor mobility in urban settings to train and enhance movement. The creators of the UrbanRehab program began with a focus group of rehabilitation specialists and aimed to create an environment where patients could be challenged with a variety of different tasks in a realistic cityscape. The challenges and achievements of the UrbanRehab program are reported from a developer’s perspective in further detail [104], highlighting the development-related challenges of designing and implementing realistic VR games for rehabilitation. For example, designing realistic games that enable functional movement is limited by in-home space constraints, presenting a common and difficult challenge that developers must overcome. However, the UrbanRehab initiative represents an exciting first step in realizing the potential for VR to reduce real-world falls.

A major advantage of VR is the capability to address individual differences due to its capacity to change the level of task difficulty and create appropriate “challenge points” or “desirable difficulties” [105]. To better detect everyday fall risk, a range of typical sensory, motor, and cognitive challenge points can be incorporated into a contextually appropriate VR-based stress test for fall risk. For instance, incorporating increasingly physiologically demanding locomotor tasks as challenge points, such as changes in walking speed along a crosswalk, could better reveal fall risk related to poor lower limb strength or power [106] and may be a better predictor of fall risk than measuring walking at comfortable speeds [107]. VR-based challenge points can present increasingly difficult but realistic obstacles for avoidance such as potholes, puddles, or oncoming pedestrians [108], which would better identify people who struggle to maintain balance during complex stepping tasks [109]. Challenging balance control in VR with dynamic locomotor tasks such as turning would reveal deficits in motor control mechanisms that are relevant to everyday falls [110,111]. Using challenge points to progressively manipulate visual flow during walking [46,57] or walking in increasingly low-light settings [112] would flag sensory integration deficits for intervention (ie, proper corrective eye wear, visuo-motor training) before they become a problem [113]. Finally, presenting increasingly complicated but realistic concurrent cognitive tasks as challenge points, such as concurrently monitoring a clock or oncoming pedestrians [99], would better identify older people at risk of real-world falls due to cognitive-motoric deficits [114,115]. In sum, VR allows scientists to measure performance across different challenge points, whereby participants are pushed to perform at, or near, the limits of their capability in everyday cognitive-motor performance, highlighting locomotor deficits before they lead to a fall in daily life.

The individual differences in learning outcomes detected across challenge points in diagnostics can optimize VR-based interventions for each learner and lead to enhanced transfer of skills to the real world. VR lends itself to quantifying and adjusting training loads for individual responses to an intervention and augmenting training loads for learners [105]. Commercially available VR systems boast multiple ways to measure performance (eg, eye tracking and kinematics) to achieve goal-related outcomes (eg, a gamified performance goal), and either aspect of a simulation can be tailored to target individual needs and training goals. Advances in data modeling and machine learning allow us to measure individual differences in learning by analyzing changes in performance versus goal-related outcomes (ie, speed versus accuracy analyses) across repeated practice sessions or challenge points. Using VR to quantify individual learning trajectories allows for tailored challenge points and maintaining appropriate difficulty levels [105], and VR-based approaches could adjust training loads to optimize learning for each participant throughout an intervention.

Practitioners are already using VR with younger populations to rehabilitate those with anxiety-related disorders, pioneering a wider application of VR for “exposure therapy” to reduce anxiety [100]. For example, the Bravemind program builds on evidence highlighting that VR-based exposure therapy can reduce depression and anxiety in veteran soldiers with PTSD [116]. Bravemind uses VR to simulate combat-related scenarios in combination with cognitive behavioral therapy, which reframes anxiety-based reactions with “cognitive restructuring,” providing a low-threat context where a patient can therapeutically decondition the cycle of anxiety via habituation [100]. The immersive VR program trains soldiers to compensate and adapt to their PTSD symptoms over time with prolonged exposure to threatening stimuli in VR, guided by evidence from successful VR treatments leveraging exposure therapy for specific phobias [69,117]. Similar exposure effects are reported in VR-based anxiety-inducing settings; the effects of mobility-related anxiety on motor performance (ie, gait speed) decay as participants acclimate to the simulated scenario [23,24], suggesting a similar approach may work to train older adults at risk for maladaptive mobility-related anxiety.

The use of VR-based approaches in conjunction with cognitive behavioral therapy could be effective in reducing fear of falling in older adults [118]. VR-based therapies for treating anxiety disorders are rising in popularity, and meta-analyses support using VR for treating anxiety when compared with traditional therapies [119], especially in preventing patient attrition [119,120]. While no substantive advantages are associated with current VR-based programs compared to traditional therapy, participants are more likely to adhere to VR-based therapies, thus improving their efficacy. We speculate that higher levels of adherence could be a result of greater engagement, which improves motor learning outcomes in gamified rehabilitation programs as compared to the sterility of traditional rehabilitation settings [64]. If VR technology can be refined and harnessed, we believe that the effectiveness (and relative affordability) of such therapies can outperform traditional treatment methods. Because of the success of programs such as Bravemind and the increasingly immersive VR experiences, it is not difficult to imagine a future where older adults are trained to optimize their performance on everyday mobility tasks using controlled exposure to stressful virtual environments.

VR is an appealing and practical tool to enhance and maintain older adult mobility. We envision a future where realistic contexts for individualized VR-based fall-prevention diagnostics and interventions will reduce the risk of real-world falls in older adults. By incorporating the context of daily tasks into VR-based approaches, we can quantify movement within complex and realistic environments that better represent the demands that older adults navigate every day. Focusing on using VR to mimic the influence of affect on cognitive, perceptual, and motor behaviors should guide industry, researchers, and clinicians toward therapies that prevent everyday falls.