The subjects included healthy individuals aged 18 to 60, excluding those who require walking aids such as canes, as well as those with neuromuscular diseases, cardiovascular diseases, respiratory diseases, or motor diseases that may influence their walking. The study was discontinued if a subject complained of pain or discomfort when using the insole. 18 males and females (9 males, 9 females, from 20 to 63 years old, average 43.9 ± 14.4 years old) satisfying the above criteria were selected.
For comparison of the insoles among subjects, we asked subjects to walk wearing standardized shoes (no heel counter and shank) with normal flat insoles (hereinafter, "flat insole") without an arched shape on the surface and made of polyurethane and wearing carbon insoles (BMZ insole, BMZ. Inc, Gunma) (hereinafter, "functional insole") (Figs.1 and Additional file 1 ) and conducted a gait analysis using a three-dimensional motion analysis device (Vicon MX, Vicon Motion Systems Oxford. UK,) and a force plate (AMTI, Watertown, MA, USA).
For the three-dimensional movement analysis, analysis markers were attached to a total of 28 points: 4 points at the head, C7, Th8 spinous processes, the midpoint of the left and right superior posterior iliac spines, the acromion, the external humerus condyle, the radial styloid process, the superior anterior iliac spine, the point 1/3 from the greater trochanter on the line between the superior anterior iliac spine and the greater trochanter, the external femoral condyle, the midpoint of the line shape between the external femoral condyle and the ankle lateral malleolus, the ankle lateral malleolus, the midline of the facies posterior to the calcaneus, and the head of the second metatarsal bone.
Patients walked normally in standardized shoes on a walkway of approximately 10 m without knowing which insole they are wearing. The insoles to be worn were randomly chosen. No instructions on walking speed were given and the subjects walked at their optimal speed. After two trials each with both the functional insoles and the flat insoles, a total of three measurements were taken.
In order to investigate any correction due to the insoles, a past study compared the control conditions wearing the shoes the subjects regularly wore with 3 mm flat insoles and wearing standardized shoes with the same insoles 11). As a result, the conclusion recommends using the footwear of the participants as the control condition, because standardized shoes, compared to the usual shoes, significantly affect the knee adduction impulse, ankle abduction moment, and vertical grounding reaction load factor at the time of knee abduction. However, since walking needs to be evaluated with "shoes + insole", this study focused on the difference in the function of insoles by using the same shoes (no heel counter and no shank).
All subjects completed the measurements without any adverse events. The obtained data was imported into the VISUAL 3D （Ver. 6） Visual 3D software program, ver. 6 (C-Motion, Inc., Germantown, MD, USA) to calculate the walking parameters and kinematics data.
Walking speed, cadence, step length, and stride length were recorded as walking parameters. Walking speed was defined as the speed of movement of the center of gravity in the direction of travel, calculated as the average of 5 m as the middle point during walking. Cadence was also calculated at 5 m as the midpoint during walking. The step length and stride length were also calculated as the average of 5 m as the midpoint during walking.
We calculated hip and ankle moment on the sagittal plane at terminal stance, and analysed them. The peak values of hip flextion moment, hip power on the sagittal plane, ankle planter flextion moment, and ankle power on the sagittal plane were recorded.
We compared the data when wearing flat insoles and that when wearing functional insoles. Group A (n = 7) included those with a left-right difference of more than 20% in ankle power when using flat insoles, while Group B (n = 11) included those with a difference of less than 20%. We compared the subjects within each group.
Regarding statistical processing, we performed a paired t-test for the walking parameters and Wilcoxon's signed rank test for left-right differences in joint moment, power, and ankle power, with <0.05 considered to be a significant difference.