Effects of sampling frequency and interpolation preprocessing on inter- and intra-study comparability of stabilometric variables

Objective: The center of pressure (CoP) is commonly assessed during static standing; however, instrumental measurement performance critically impacts this variable. We focused on sampling frequency and systematically identified the effects of sampling frequency and interpolation preprocessing on inter- and intra-study comparability of stabilometric variables.Methods: Stabilometric variables in four domains-positional, dynamic, frequency, and stochastic-were investigated. CoP data were collected at 1000 Hz, and the variables obtained from data downsampled to 100 Hz were treated as "real values." Additionally, variables were obtained from low-sampling frequency measurement data, which were downsampled to 50 Hz and 20 Hz and upsampled to 100 Hz through interpolation applied to simulated 50-Hz and 20-Hz data. The inter-study comparability was assessed via the effect size and minimal detectable change with 95% certainty, and the intra-study comparability was assessed via the coefficient of determination.Results: Although a sampling frequency of 20 Hz affected the intra-study comparability of one dynamic variable and four frequency variables, interpolation preprocessing restored the intra-study comparability of all variables. Moreover, low-sampling frequency conditions affected the inter-study comparability, except for positional variables. Interpolation preprocessing restored the inter-study comparability of most variables; however, the number of crossings of the CoP velocity in the interpolated 50-Hz and 20-Hz data as well as the short-term diffusion and scaling coefficients in the interpolated 20 Hz data were still affected.Conclusion: Although some variables require careful handling, upsampling improves the consistency of variables and simplifies the comparison and interpretation of the results, including low-sampling frequency measurement data.