Test-retest reliability of an insole plantar pressure system to assess gait along linear and curved trajectories

BACKGROUND: Previous studies have assessed reliability of insole technology for evaluating foot pressure distribution during linear walking. Since in natural motion straight walking is intermingled with turns, we determined the test-retest reliability of insole assessment for curved as well as linear trajectories, and estimated the minimum number of steps required to obtain excellent reliability for each output variable. METHODS: Sixteen young healthy participants were recruited. Each performed, two days apart, two sessions of three walking conditions: linear (LIN) and curved, clockwise (CW) and counter-clockwise (CCW). The Pedar-X system was used to collect pressure distribution. Foot print was analyzed both as a whole and as subdivided into eight regions: medial and lateral heel, medial and lateral arch, I metatarsal head, II-V metatarsal heads, hallux, lateral toes. Reliability was assessed by using intraclass correlation coefficient (ICC) for clinically relevant variables from analysis of 50 steps per trajectory: Peak Force (PF); Peak Pressure (PP); Contact Area (CA); Stance Duration (S). RESULTS: When considering whole-foot, all variables showed an ICC >0.80, therefore highly reliable. This was true for both LIN and curved trajectories. There was no difference in ICC of the four variables between left and right foot. When collapsing foot and trajectories, S had a lower ICC than PP and CA, and PP lower than CA. Mean percent error between the values of first and second session was <5%. When separately considering the eight foot regions, ICCs of PF, PP and CA for all regions and trajectories were generally >0.90, indicating excellent reliability. In curved trajectories, S showed smaller ICCs. Since the least ICC value for S was 0.60 in LIN trajectory, we estimated that to achieve an ICC ≥0.90 more than 200 steps should be collected. CONCLUSIONS: High reliability of insole dynamic variables (PF, PP, CA) is obtained with 50 steps using the Pedar-X system. On the contrary, high reliability of temporal variable (S) requires a larger step number. The negligible differences in ICC between LIN and curved trajectory allow use of this device for gait assessment along mixed trajectories in both clinical and research setting.