Study at a glance
Summary
Disclaimer: The following summary details independent academic research. Theia3D is intended for research, sports performance and educational use only. It is not cleared as a medical device and is not intended to be used for the diagnosis, treatment, mitigation or prevention of any disease, injury or medical condition.
Editor's note: Theia3D version v2023.1.0.3161, as stated in the paper. Here, Theia3D served as the in-person assessment tool generating one of three mobility domains examined, rather than as a system being validated against marker-based motion capture within this particular study.
Study Overview
Design
A cross-sectional cluster analysis combining self-reported function, in-person markerless motion capture assessment, and seven-day free-living wearable-sensor data to identify pre-surgical mobility profiles in adults awaiting knee replacement, with a secondary analysis examining how consistently individuals ranked across these three domains.
Participants
Fifty-six adults with knee osteoarthritis (moderate to severe, confirmed by imaging or assessment at the time of surgery) recruited approximately two weeks before a scheduled knee replacement procedure, able to walk independently and provide informed consent.
Data Collection
Participants completed self-reported questionnaires covering knee-specific function and pain, general quality of life, and depressive symptoms. They then completed an in-person assessment of functional tasks, a 60-second comfortable-pace walk, and, in a subset of 37 participants, a 30-second faster-paced walk and a five-repetition sit-to-stand task, captured using a 10-camera markerless system processed through Theia3D. Before the in-person assessment, participants were fitted with a small wearable sensor on each shin to continuously record free-living walking, step counts, and sedentary time for up to seven days.
Analysis
Theia3D video data were processed to generate three-dimensional joint kinematics, with principal component analysis used to summarize key patterns in knee flexion and side-to-side knee movement across the walking cycle. Hierarchical cluster analysis grouped participants based on a combined profile of self-reported, in-person, and free-living measures, testing two models: one including all 56 participants using only the comfortable-pace walk, and a second including the 37 participants who completed every in-person task. A separate, exploratory analysis examined how consistently participants within the larger, higher-functioning cluster ranked in the same relative tertile (low, middle, or high) across the self-reported, in-person, and free-living domains.
Key Findings
- Both clustering models consistently identified two mobility profiles: a smaller, lower-functioning group and a larger, higher-functioning group, with the two groups differing significantly across self-reported function, in-person walking performance, and free-living activity levels. The lower-functioning group had a higher body mass index, a greater proportion of women, worse self-reported function and quality of life, slower in-person walking speed, and fewer daily steps with more time spent sedentary in daily life.
- The lower-functioning group also showed distinct movement pattern differences during Theia3D-based gait assessment, including reduced knee flexion magnitude and range of motion and a pattern resembling increased sideways knee loading (a "varus thrust" pattern) during walking, consistent with patterns previously linked to knee pain and stiffness in other knee osteoarthritis research.
- Within the larger, higher-functioning group, only about 11 to 15 percent of participants ranked at a consistent relative level, low, middle, or high, across the self-reported, in-person, and free-living domains, indicating that how someone performs during a brief in-person assessment often does not predict how they report feeling or how active they actually are in daily life.
- Between-group walking speed differences (roughly 0.3 meters per second at a comfortable pace) met or exceeded commonly cited thresholds for practically meaningful change, reinforcing that the two identified profiles reflect functionally distinct levels of mobility impairment rather than a statistical artifact of the clustering method.
- The findings held even though self-reported function, in-person performance, and free-living activity are sometimes assumed to move together: the authors note this disagreement with some prior domain-specific research likely reflects their study's integration of all three domains within a single combined analysis, rather than comparing domains two at a time.
What This Means for Movement Science Researchers and Pre-Surgical Mobility Assessment Teams
For researchers and teams working on pre-surgical mobility assessment in knee osteoarthritis, this study offers a clear, practical takeaway: no single measure, self-report, a brief in-person walking assessment, or free-living activity tracking, fully captures someone's overall mobility picture on its own, and relying on just one risks missing important, individual-specific patterns. The cluster-based approach used here, built in part on Theia3D-derived gait data, offers a concrete illustration of how combining markerless motion capture with complementary self-report and wearable-sensor measures can identify distinct mobility profiles and surface where an individual's function, perception, and daily activity diverge, information that could help tailor pre-surgical planning, whether that means prioritizing activity-focused strategies for someone who performs well in-person but remains inactive at home, or targeting walking-focused work for someone whose in-person capacity itself is limited.
The Role of Theia3D in Movement Analysis
Theia3D served as the in-person, markerless motion capture tool generating one of three mobility domains examined in this study, joint-level walking and sit-to-stand kinematics, alongside self-reported questionnaires and free-living wearable-sensor data; it was not itself compared against marker-based motion capture within this particular study. The authors cite prior Theia3D validation work, including concurrent agreement with marker-based systems and consistency across different assessment settings, as part of the basis for using it here. This reflects a practical use case for markerless capture: enabling efficient, joint-level movement assessment as one input within a broader, multi-domain mobility evaluation rather than requiring a dedicated motion-capture laboratory visit on its own.
Study Limitations and Future Directions
The authors note that their recommended sample size guidance for cluster analysis was met for the full 56-participant model but fell short for the 37-participant model that included the additional faster-paced walking and sit-to-stand tasks, though the same two-group pattern held in both. The cross-sectional design and focus on pre-surgical participants limit how well these specific mobility profiles generalize to other severities of knee osteoarthritis or to post-surgical populations, and the higher proportion of women and higher body mass index in the lower-functioning group may partly explain some of the between-group differences observed. The authors identify testing whether these pre-surgical mobility profiles predict how someone's function, performance, and daily activity change after surgery, and whether addressing a specific lagging domain through targeted pre-surgical preparation improves outcomes, as important next steps.
Citation: Di Bacco, V. E., Ruder, M. C., Madden, K., Adili, A., & Kobsar, D. (2026). Functional mobility profiles in pre-operative knee osteoarthritis patients: A cluster analysis of self-report, in-clinic, and free-living measures. Clinical Biomechanics, 136, 106821. https://doi.org/10.1016/j.clinbiomech.2026.106821
Link: Read the full study
Does One Mobility Measure Tell the Whole Story Before Knee Replacement Surgery?
This study of 56 adults awaiting knee replacement surgery finds that combining Theia3D-based gait assessment with self-reported function and free-living activity tracking reveals distinct mobility profiles, and that even within a higher-functioning group, in-person performance often does not match daily-life activity. If your work involves pre-surgical mobility assessment and you want to explore how markerless motion capture can fit into a multi-domain evaluation approach, talk to our team about applying Theia3D to your own research.