Familiarization: Before the first session, athletes were introduced to the headset through a brief tactile and visual familiarization period. This ensured comfort with the device and reduced anxiety for VR-naive users.

Pre-VR Emotional Assessment: Athletes self-reported their emotional state using a three-option emoji scale (Good-green, Neutral-yellow, and Bad-red). This accessible format included visual symbols, color coded, and simplified verbal prompts.

Pre-VR Balance Assessment: Two balance tasks were administered and recorded:

Two balance tasks were administered and video recorded. Static and dynamic balance were evaluated using structured observational tasks scored within a feasibility-oriented framework developed for this adapted sports context.

Static balance was assessed during quiet standing with feet close together, arms extended laterally, and eyes open for approximately 20–30 s. Dynamic balance was assessed using a heel-to-toe walking task along a straight 2-meter line, performed at a self-selected pace. All trials were video recorded to allow detailed post-hoc analysis.

Balance performance was evaluated using a two-step scoring procedure. First, predefined observable indicators were rated and summed to generate raw balance scores. For static balance, indicators included sway amplitude, foot stability, arm compensation, and postural alignment, yielding a raw score range of 0–12, where lower scores indicated better stability (0–3 = excellent stability; 4–6 = adequate stability; 7–9 = unstable; 10–12 = poor stability or loss of control). These summed scores were then mapped onto a 0–3 ordinal scale, where 0 represented stable, well-controlled performance (raw score 0–3), 1 indicated minor instability with occasional postural corrections (4–6), 2 indicated moderate instability with frequent compensatory movements (7–9), and 3 indicated pronounced instability, such as marked sway, step deviations, or need for external support (10–12).

For dynamic balance, indicators included step symmetry, trunk control, direction-change control, foot placement accuracy, and gait flow/coordination, yielding a raw score range of 0–15, where lower scores reflected better coordination (0–4 = excellent coordination; 5–8 = adequate balance; 9–12 = reduced dynamic stability; 13–15 = unstable or increased risk of fall). These summed scores were likewise mapped onto the same 0–3 ordinal scale, with thresholds adjusted accordingly (0–4 = score 0; 5–8 = score 1; 9–12 = score 2; 13–15 = score 3).

This scoring framework was not intended as a validated clinical balance assessment but as a structured, feasibility-oriented method for detecting short-term postural changes before and after VR exposure in athletes with intellectual disabilities.

To ensure scoring consistency, at least two trained observers completed a calibration process prior to data collection, including joint review of example recordings and discussion of scoring anchors. During the study, 25% of recordings were independently double-scored, yielding 90% exact agreement across parallel ratings. Any discrepancies were resolved through consensus review, ensuring stable and reliable scoring within an applied sport setting.

Although raw scores were additionally mapped onto a 0–3 ordinal severity scale for descriptive classification, all analyses and reported results are based on summed raw balance scores to preserve sensitivity to change. Headset Fitting and Calibration: The headset was adjusted individually for interpupillary distance, strap tension, visual clarity, comfort, and stability.

Researchers verified clarity using simple Yes/No questions and observed facial cues. Casting ensured continuous monitoring of the athlete's field of view, enabling rapid correction of misalignment, tracking errors, or user confusion.

The duration of VR exposure was limited to 7–10 min per session to prioritize feasibility, tolerability, and participant safety during initial immersive exposure. Short exposure durations have been recommended in exploratory VR studies and feasibility research to reduce the risk of sensory overload, attentional fatigue, and VR-induced discomfort, particularly among users with limited prior VR experience or heightened sensory sensitivity (5, 11). Studies involving individuals with intellectual disabilities further emphasize the importance of brief, structured exposure and gradual familiarization to support emotional comfort and sustained engagement (1, 2). Given the feasibility-oriented nature of the present study, longer exposure durations were intentionally avoided to allow close monitoring of emotional, behavioral, and postural responses and to minimize cumulative effects across repeated sessions (15, 17).

Trained observers documented real-time behaviors across five domains: engagement, distraction episodes, hesitation events, verbal expressions, and physical cues.

The behavioral coding framework was adapted from established VR engagement literature and observational models used in adapted physical activity, ensuring that the selected categories reflect meaningful indicators of attention, comfort, and cognitive load in individuals with ID (2, 6, 8, 15, 17). Each domain was operationalized using predefined behavioral descriptors and frequency-based scoring anchors. Behavioral indicators were scored using frequency-based anchors defined a priori for each domain.

Engagement was scored based on sustained visual orientation toward the VR content and spontaneous interest behaviors (e.g., verbal comments, focused posture), with higher scores indicating longer and more consistent engagement. Distraction was coded as observable shifts of attention away from the VR content (e.g., looking away, interacting with the environment), with higher scores reflecting more frequent episodes. Hesitation was defined as pauses, uncertainty, or delayed responses during VR exposure. Verbal expressions included spontaneous comments, exclamations, or questions related to the VR content, and physical cues captured observable bodily reactions such as gestures, posture adjustments, or expressive movements. All behaviors were coded using predefined frequency thresholds to ensure consistency across observers and sessions. Observers completed joint calibration sessions prior to data collection to ensure shared interpretation and consistent scoring.

During VR exposure, one of the observers was positioned in close proximity to the athlete (approximately 1 m) to ensure immediate physical support if needed, particularly in cases where an athlete attempted to stand, move, or interact with the environment while wearing the headset. This positioning was selected to minimize the risk of falls or collisions during immersive exposure and to allow rapid intervention in the event of unexpected movements.

Monitoring for Cyber sickness and Discomfort: Using SSQ-aligned observational criteria, staff monitored early signs of VR-induced discomfort, including: eye rubbing, excessive blinking, or gaze aversion, postural instability, sudden stepping, or bracing actions, facial tension, pallor, or other indicators of nausea, behaviors suggestive of disorientation or sensory overload.

Athletes were permitted to stop the session at any time.

Usability and Technical Logging: All technical issues, including tracking loss, video orientation errors, accidental menu activation, battery depletion, overheating, and controller-related disruptions, were systematically logged.

User-related challenges, such as difficulty maintaining head alignment, interpreting prompts, or managing sensory flow, were recorded in parallel to support feasibility evaluation.

Emotional State (Post-VR): The same emoji-based scale used during the pre-session emotional assessment was used to capture immediate emotional responses after headset removal.

Post-VR Balance Assessment: The same static and dynamic balance tasks as the pre-session were repeated to detect short-term VR-related changes.

A short, structured debrief was conducted using simple, concrete prompts supported by gestures and demonstrations to facilitate understanding. Prompts focused on immediately observable sensations and included questions such as: “Do you feel dizzy now?”, “Do your eyes feel tired?”, “Do you feel unsteady when standing?”, and “Do you want to stop or rest?”. Verbal prompts were accompanied by pointing to the head, eyes, or body, and by demonstrating standing still or taking a step, to support comprehension.

Session Repetition: The three VR sessions were conducted on separate training days with a consistent interval of six days between sessions for all participants. All sessions took place at the same time of day for each athlete to minimize potential effects of fatigue, circadian variation, or training schedule differences.

Balance assessments conducted immediately before and after each VR, exposure revealed no evidence of postural deterioration following immersive VR viewing. Across sessions, athletes demonstrated either stable or improved performance in both static and dynamic balance tasks.

At baseline, athletes displayed heterogeneous balance profiles, consistent with typical inter-individual variability among persons with intellectual disabilities. Four athletes (Athletes 2, 3, 5, and 6) demonstrated relatively stable posture and coordinated gait, characterized by minimal sway and accurate step placement. In contrast, two athletes (Athletes 1 and 4) exhibited moderate instability, including observable postural sway, compensatory trunk movements, and deviations during heel-to-toe walking. Group-level mean raw balance scores reflected this variability, with mean summed scores of 4.0 for static balance (0–12 scale) and 4.7 for dynamic balance (0–15 scale).

Following VR exposure, no athlete demonstrated a decline in either static or dynamic balance performance. Instead, all participants either maintained or improved their balance scores Mean raw post-VR scores decreased to 1.3 for both static (0–12 scale) and dynamic (0–15 scale) balance, indicating greater postural stability (lower scores reflect better performance; Table 3).

Values are summed raw observational scores, lower scores indicate better balance and Δ = Post - Pre (negative values indicate improvement).

Notably, four athletes achieved optimal post-VR scores (score = 0) on both balance tasks. Athletes who exhibited the weakest baseline stability (Athletes 1 and 4) showed improvements, with score reductions of up to four points.

Across all post-exposure assessments, no increases in sway, missteps, disorientation, or signs of vestibular-visual conflict were observed. No athlete required additional support, rest, or recovery time following VR exposure. Overall, these findings indicate that exposure to low-motion, non-interactive 360° VR content did not negatively affect motor stability and was physiologically well tolerated by athletes with intellectual disabilities.

All athletes tolerated the Meta Quest 3 well. No adjustments were needed during sessions, and no discomfort related to strap tension, headset weight, or eye alignment was observed. These findings indicate that the device's ergonomic design was appropriate for this population.

No athlete showed signs of blurred vision or IPD misalignment, and recalibration was never required. The main visual challenge involved 360° video orientation: when videos started facing a different direction than the athlete, assistance was needed to orient to the correct viewpoint. Video initiation also proved sensitive to headset transfer, often causing pauses or resets. Because most athletes could not restart content by themselves, researchers intervened frequently. Continuous casting enabled rapid detection and correction of these disruptions.

Controller use was a significant usability barrier. Only two athletes demonstrated purposeful controller interaction; others triggered unintended functions through hand movements, including accidental menu activation and video interruption. In some cases, sessions needed to be restarted. When used correctly, pointing and trigger activation were performed without difficulty, though this was limited to a small subset of participants.

Casting the VR view to an external monitor functioned reliably, with only brief lags that did not affect safety or session flow. Instructors maintained continuous visual access to the athlete's perspective, supporting timely troubleshooting.

Several athletes experienced difficulty interpreting on-screen prompts and required consistent verbal scaffolding. Despite this, transitions between virtual and real-world tasks remained smooth, and no confusion emerged during headset removal or repositioning.

No athlete demonstrated resistance to the headset, heightened arousal, or behavioral withdrawal. Sessions proceeded without the need for breaks or emotional redirection, indicating high acceptance of VR.

The session environment supported stable tracking and presented minimal distractions. No modifications to lighting, space, or layout were required.

Logistical challenges included extended setup time and highly challenging cleaning and disinfection, occasional technical interruptions, rapid battery depletion of both headset and controllers, and intermittent device overheating. These factors increased staff workload and should be anticipated when planning VR use with multiple athletes.