Physical Medicine

Knee OA exercise app boosts adherence without improving function

Adults with knee osteoarthritis and functional limitations performed physiotherapist-prescribed strengthening three times weekly, with or without the My Exercise Messages app. At 26 weeks, the app group had higher complete adherence to prescribed frequency (36% vs 16%; relative risk 2.1, p<0.001). Despite better adherence, WOMAC function scores did not differ between groups at 26 weeks (mean difference -0.9 on 0–68 scale, p=0.52). Other secondary outcomes, including pain, sport and recreation function, quality of life, and physical activity, were also similar between groups. 1

Supervised rehab offers no advantage over single advice session after proximal humerus fracture

Older adults with displaced proximal humerus fractures treated non-operatively all received one structured advice session on gradual activity resumption. They were then randomized to municipal supervised rehabilitation or no further formal training, with routine radiographic follow-up for both groups. At 6 months, Oxford Shoulder Score was nearly identical between groups (33.9 vs 33.8; mean difference 0.05, 95% CI -4.5 to 4.6). Secondary outcomes, including 12‑month function, quality of life, conversion to surgery, and adverse events, were consistent with no clinically relevant benefit. 2

Mobility-Fit improves upper-limb strength and quality of life in long-term care

Twenty long-term care facilities were cluster-randomized to Mobility-Fit, a 12‑week multicomponent program, or standard care with lower-limb exercises. Participants (median age 86, 60% female) attended tailored 45‑minute sessions three times weekly with high adherence (83.3%). Mobility-Fit produced superior gains in elbow extension strength and health-related quality of life (EQ-5D utility and visual analogue scores). Both groups improved knee extension strength, sit-to-stand performance, and fall-risk profile, reflecting benefits of structured exercise overall. 3

Markerless 3D video gait analysis enables scalable kinematics in spinal cord injury

Researchers benchmarked multiple 3D pose-estimation models and identified VideoPose3D as best performing for pathological gait in spinal cord injury. Their automated pipeline uses a single video to generate time-series joint angle data and detailed kinematic profiles. In 225 individuals with spinal cord injury, the system automatically distinguished gait patterns, including reduced hip and knee flexion at specific gait phases. This markerless, non-invasive approach enables remote, real-world gait assessment without motion labs or reflective markers. 4