Variability and stability indexes in Dual-Task condition: Methodological and physiological aspects

Introduction: Instrumented gait assessment is traditionally carried out asking subjects to merely walk (ST). Such condition does not represent an everyday walking situation; indeed, in the daily life, locomotion requires to meet individual goals, to adapt to the environmental constrains and commonly involves cognitive tasks. Dual-task (DT) paradigms have been used to assess a shift in gait control strategy in real life situations [1]. In particular, variability and stability indexes used to motor control assessment, could be useful to evaluate cognitive motor interference during walking. The aim of the present study is to evaluate how the DT condition changes the stability and variability indexes. Methods: 16 young adults (25YA) participated to the study. Measures of trunk acceleration (L5 level) were collected using one tri-axial wireless inertial sensor. The participants were asked to walk for 250 m: at self-selected speed (ST) and counting backward aloud by 7 from a random starting number (DT). For all the participants and each condition 5 variability indexes – Standard Deviation (SD), Coefficient of Variation (CV), Non-stationary Index (NI), Inconsistency of Variance (IV) and Poincaré Plots (PSD1/PSD2) – and 6 stability indexes – short/long-term Lyapunov Exponents (sLE/lLE), Harmonic Ratio (HR), Index of Harmonicity (IH), Multiscale Entropy (MSE) and Recurrence Quantification Analysis (RQA) – were calculated on non-overlapping windows of 20 strides and on the whole signal. Kruskal–Wallis tests (p-value at 5%) were performed to evaluate DT effects. Results: Analyses performed on the 6 portions of 20 strides showed, in general, three different behaviours: none/all the portions were statistically significant or only the first portions were significantly different. In particular, for HR, was found that the first two portions showed statistical differences. Kruskal–Wallis test performed between indexes calculated on the entire signal did not show statistical differences (Fig. 1). Discussion: Analyses on the six portions of 20 strides suggest that indexes were influenced by DT condition; furthermore, indicate the existence of an adaptation, by the subjects, to the DT conditions. The statistical results about the indexes calculated on the entire signal were not in agreement with those in literature [2,3]. A non-sufficient analyzed strides number, taken into account in the previous studies, to guarantee that the indexes reached steady values [4] could explain these discrepancies. In conclusion the obtained results highlight the importance of a standardized methodology to compute indexes and suggest that DT condition during trial seems to fit with the walking pattern of the ST. Further researches are needed to understand the physiological and motor control meaning of the adaptation phases.