Purpose: Patellar bracing is a common, mechanical-based treatment strategy for patellofemoral osteoarthritis (OA). It is thought that the brace corrects patellar tracking, however, this correction has not been quantified in the OA population. Through advances in magnetic resonance imaging (MRI), we can now assess patellar tracking in three-dimensions.

Method: We assessed three-dimensional patellar tracking in ten subjects with symptomatic radiographic patellofemoral knee OA using a validated, quasi-static, MRI-based method. Four conditions were studied:

no knee brace, no load,

Patellar tracking (flexion, spin and tilt; proximal, lateral and anterior translation) was assessed. Comparisons were made at 1° increments over the coincidental range of knee flexion between the no-brace and brace conditions, at no load and 15% BW load, using a paired t-test with Bonferroni correction.

Results: All subjects (7 female, 3 male, 60.9±1.3 yrs, 89.5±19.3 kg) had radiographic lateral patellofemoral OA and seven had concomitant tibiofemoral OA (KL grade≥2). Under no load, the brace extended (mean=2.7°, CI=[2.4°, 2.9°], P< 0.001) and medially tilted (mean=−1.4°, CI=[−1.6°, −1.2°], P< 0.001) the patellae and shifted them distally (mean=0.8mm, CI=[0.6mm, 0.9mm], P< 0.001), medially (mean=0.5mm, CI=[0.5mm, 0.6mm], P< 0.001) and posteriorly (mean=0.6mm, CI=[0.5mm, 0.6mm], P< 0.001). Under 15% BW load, the brace extended the patella (mean=2.4°, CI=[2.1°, 2.8°], P< 0.001) and shifted them distally (mean=1.3mm, CI=[1.1mm, 1.4mm], P< 0.001), medially (mean=0.8mm, CI=[0.7mm, 0.9mm], P< 0.001) and posteriorly (mean 0.6mm, CI=[0.5mm, 0.7mm], P< 0.001).

Conclusion: The brace extended the patellae for both loading conditions, suggesting that patellar flexion/extension is restricted by the brace. The brace tilted the patellae medially under no load only, suggesting when the quadriceps are active (15% BW load) the brace has little effect for tilt. While the effect of bracing on patellar tracking may appear small, the differences are of similar magnitude to those observed between normals and patients with patellofemoral pain, suggesting that braces may produce clinically significant changes in patellar tracking.