Reliability and Validity of Augmented Reality-Based Clinical Gait and Balance Assessments in Healthy Community-Dwelling Older Adults

Falling is a rapidly growing, costly and often impactful public-health problem in our ageing society (Lord et al., 2001; Rubenstein et al., 2006; Milisen et al., 2017). Due to shortage of healthcare professionals, new ideas and approaches are needed to reduce falls beyond the state-of-the-art of fall prevention that can be implemented independently at home. Augmented reality (AR) based assessments may provide an efficient, engaging and standardised way to evaluate gait and balance. This study is used to test reliability and validity of AR technology in assessing fall risk of healthy older adults. AR has already demonstrated good reliability and validity in measuring standard clinical gait and balance tests in people with Parkinson's disease, but this has not yet been established in healthy community-dwelling older adults (van Bergem et al., 2024; van Doorn et al., 2025). Confirming test-retest reliability, criterion and known-groups validity of AR-based assessments in this population is an essential first step towards a safe and effective integration of AR into the assessment and treatment of fall risk. This study therefore aims to investigate a set of standardised clinical gait and balance assessments that have been implemented in an AR application and are performed using the state-of-the-art Magic Leap 2 AR glasses. Purpose of the study is to evaluate key elements of reliability and validity of these AR-based assessments compared to the same tests administered in the conventional form, i.e. without the use of AR technology. We expect that the AR-based clinical assessments will show good to excellent within-day test-retest reliability, with reliability results comparable to those of the conventional clinical assessments of fall risk indicators. Furthermore, we expect AR-based clinical assessments to show strong correlations and good agreement with the corresponding gold-standard clinical tests performed without the AR glasses. Lastly, we expect that both AR-based and conventional forms of clinical assessments can distinguish between individuals with and without a history of falls. Ultimately, the study may serve as a foundation for subsequent development of more AR-implemented assessments that are intended to assess abilities required to avoid walking-related falls.