gms | German Medical Science

Objectives: Analyses of gait patterns via insoles are increasingly used to monitor treatment progress. Despite the popularity of pedography, characteristic effects of individual parameters on the M-shaped curve trajectory have not been studied. Walking uphill or downhill could also affect the curve in typical ways. We hypothesized that age, height, body weight, body mass index (BMI), and handgrip strength, as well as walking on a slope have a characteristic influence on the plantar pressure curve in the stance phase.

Methods: 37 participants (18 women, 19 men) with an average age of 43.65 ± 17.59 y were fitted with Moticon OpenGO insoles equipped with 16 pressure sensors each. Data were recorded from both feet at a frequency of 100 Hz while the participants walked for one minute at 4 km/h on a level treadmill. In addition, data were recorded with the same setup from 40 healthy participants of both sexes (19 women and 21 men, average age 43.90 ± 17.30 y) in each of the following slopes: -20, -15, -10, -5, 0, 5, 10, 15, 20%. Data were processed via a custom-developed data platform, including step detection and parameter calculation. Characteristic correlations of the pressure curve parameters with age, body height, body weight, body mass index and handgrip strength were assessed via multiple linear regression analysis. ANOVA analysis was conducted with gait parameters as dependent and slope as independent variables.

Results and conclusion: Age showed a negative correlation with mean loading slope (P=.014). Body height correlated with mean loading force and loading slope (P=.046, P=.023), body weight with all parameters, except loading slope (Fmstance P<.001, Fmload P=.002, Fmmid P=.002, Fmunload P<.001, Fz2 P=.002, Fz3 P=.007, Fz4 P<.001, unloading slope P=.002), BMI with all parameters, except loading slope (Fmstance P<.001, Fmload P=.005, Fmmid P=.002, Fmunload P=.002, Fz2 P=.003, Fz3 P=.008, Fz4 P<.001, unloading slope P=.002), and handgrip strength correlated with changes in the second half of the stance phase (Fmstance P=.015, Fmmid P=.036, Fmunload P=.012, Fz3 P=0.041, Fz4 P=0.015, unloading slope P=0.032). Up to 46% of the variability can be explained by these parameters. Uphill and downhill walking was associated with changes in mean force during loading and unloading, correlated with the two maxima and the minimum, and the loading and unloading slope (all P<.001). A simultaneous increase in loading slope, first maximum and mean loading force combined with a decrease in mean unloading force, second maximum, and unloading slope indicates downhill walking. The opposite indicates uphill walking. The minimum has its peak at horizontal walking and values drop when walking uphill and downhill alike. It is therefore not a suitable parameter to distinguish between uphill and downhill walking. While patient-related factors shape the stance phase curve on a longer-term scale in typical ways, walking on slopes leads to temporary characteristic short-term changes in the curve trajectory.