Predicting Dynamic Foot Function From Static Foot Posture: Comparison Between Visual Assessment, Motion Analysis, and a Commercially Available Depth Camera

• Published in: The journal of orthopaedic and sports physical therapy Vol. 45; no. 10; pp. 789 - 798
• Main Authors: Paterson, Kade L, Clark, Ross A, Mullins, Alexandra, Bryant, Adam L, Mentiplay, Benjamin F
• Format: Journal Article
• Language: English
• Published: United States 01-10-2015
• Subjects: Biomechanical Phenomena; Foot - physiology; Gait - physiology; Humans; Male; Physiology - methods; Posture - physiology; Time and Motion Studies; Walking - physiology; Young Adult; gaming; biomechanics; gait; posture; FPI; kinematic
• DOI: 10.2519/jospt.2015.5616
• EISSN: 1938-1344
• ISSN: 0190-6011

STUDY DESIGN :Controlled laboratory study.
To evaluate the ability of 3 methods to assess static foot posture to predict rearfoot and midfoot kinematics during gait.
Static foot posture is commonly used clinically to infer dynamic function. Limitations of static clinical assessments may be overcome through advances in technologies, including commercially available depth cameras.
The Foot Posture Index (FPI) of 31 males (average age, 22.5 years) was assessed using visual observation, a 3-D motion-analysis system, and a depth camera. Pearson correlations were used to evaluate relationships between FPI items and rearfoot and midfoot kinematics during walking. The ability of the static variables to predict dynamic function was assessed using multiple linear regression.
Most FPI items (85%) were not correlated with foot kinematics, regardless of assessment method. There were 6 fair to moderate correlations between visual FPI items and total rearfoot (r = -0.36 to -0.39, P<.05) and midfoot (r = 0.37 to 0.61, P<.05) motion, 2 fair correlations between 3-D motion-analysis FPI items and total midfoot (r = -0.43, P = .02) and peak rearfoot (r = -0.40, P = .03) motion, and 2 fair correlations between the depth-camera FPI items and average rearfoot (r = -0.38 to 0.44, P<.05) motion. Visual assessment of the FPI provided the best prediction model, explaining 37% of the variance in total midfoot inversion/eversion.
Static measures of foot posture are weakly correlated with rearfoot or midfoot kinematics, and have limited dynamic prediction ability. Our findings suggest that the FPI may not be an accurate representation of rearfoot or midfoot movement during walking, regardless of the measurement technique employed.