The three-dimensional kinematics of the tibio-femoral and patellofemoral joints were examined during walking. Steinmann traction pins were inserted into the cortices of the femur, the pate.lla and the tib ia of five symptom-free subjects. The subjects walked three times with four differen t types of footwear:​ barefoot, regular running shoes, running shoes with a 10 degree varus wedge and running shoes with a 10 degree valgus wedge. Each trial was filmed simultaneously by four high-speed cameras operating at 100 frames per second. Three-dimensional spatial coordinates of the target markers were determined using cinematographical photogrammetry and three-dimensional anatomical coordinates of the target markers were determined using X-ray photogrammetry. The transformation matrices required to express the location of target markers from one frame of reference to another frame of reference were computed using an exact method combined with kinematic vectors.

Six components of motion for both tibio-femoral joint (TFJ) and patello-femoral joint (PFJ) were determined for the whole duration of the gait cycle when the subjects walked with regular shoes. The spatial orientation of the tib ia and femur were also computed during the gait cycle. The angular kinematics of the TFJ and PFJ were resolved into components around the axes of specific joint coordinate systems:​ two angular components were measured around axes fixed to the bones and one perpendicular to the two bone fixed axes. The translational components of motion were measured along these same axes. The changes of kinematic pattern brought about by the different types of footwear were investigated during the stance phase of the gait cycle with the use of a trend approach.

The results have indicated that intra-cortical pins were a successful and accurate technique to measure the kinematics of the TFJ, but less successful to measure the kinematics of the PFJ due to poor pin fixation into the patella. I t was found that, in general, the subjects exhibited fa irly sim ilar patterns of motion of the TFJ and PFJ during the gait cycle, and that good relationships existed between the linear motions of the tib ia with respect to the femur and the flexion-extension of the TFJ. Similar good relationships existed between the angular and linear motions of the PFJ and the flexion-extension of the TFJ. A major finding of this study suggested that the "screw home" mechanism of the TFJ did not take place during the walking gait cycle.

Concerning the different conditions of footwear, the results have indicated that they induced numerous changes in the kinematic patterns of the TFJ and PFJ. Yet, those changes were relatively small considering the radical differences that existed between the footwear conditions and, in general, the linear changes were more important than the angular changes. During the middle part of the stance phase, the valgus shoe condition caused the tib ia to rotate internally 3.5 degrees more than the varus shoe condition, while the rotation of the TFJ was sim ilar under both conditions of footwear. The valgus shoe condition also increased the medial tibial shift by 1.5 millimeters during the whole stance phase as compared to the varus shoe condition. However, both shoe conditions led to similar motion of the patella along the medio-lateral axis of the femur. When compared to the other conditions of footwear, the regular shoes limited the contact of the femur to a more constant portion of the tibial plateau in the frontal plane. Finally, the barefoot condition induced less motion of the patella in the sagittal than the other footwear conditions.

In view of the successes obtained in podiatric practice with the use of in-shoe orthotic devices, the present results on the effects of the footwear conditions imply that the knee joints (TFJ and PFJ) are highly sensitive to minor in trin sic derangements caused by the footwear conditions.