Causal relationship between spatiotemporal parameters of walking and locomotor rehabilitation index in healthy people

Published: 11 March 2021| Version 1 | DOI: 10.17632/fsgsymgw7p.1
Alberito Rodrigo Carvalho


To identify the causal relationship between the stride length-frequency walking ratio index (R_SL/SF), both in preferred self-selected speed (PSWS) and in the optimal walking speed (OWS), with Locomotor Rehabilitation Index (LRI) in young adults and healthy adults.


Steps to reproduce

For causal relationships, we applied multiple linear regression using the forced regression method. Data multicollinearity was tested by Variance Inflation Factor (VIF) and by tolerance statistic, which is 1/IVF. Values for VIF less than 10 and tolerance statistics greater than 0.20 indicate that multicollinearity does not affect statistical analysis. We used the Pearson Correlation to test whether there is a statistically significant linear relationship between outcomes. For the comparisons between the spatiotemporal parameters in the PSWS and OWS, the statistical test used was the generalized estimation equation (GEE) model with data analysis based on the intention to treat using the maximum likelihood principle for extrapolation of the missing data. The best fit of the data was tested by two distribution models: Linear and Gamma. The model that obtained the lowest value of quasi-likelihood under independence model criterion (QIC), was chosen as the model with the best fit. The Sidak test was used as a post hoc test.


Universidade Estadual do Oeste do Parana


Gait, Walking Speed, Locomotor Rehabilitation