Reliability of recurrence quantification analysis of postural sway data. A comparison of two methods to determine recurrence thresholds

Abstract

Ageing affects balance control resulting in a greater amplitude of sway and alterations in structure of the sway time series. Recurrence quantification analysis (RQA) has been used to determine the structure of center-of-pressure (CoP; a measure that reflects standing postural control) data as a means to quantify how CoP repeats itself / recurs below a certain threshold. This study aimed to determine how the method of threshold determination, below which a recurrence is defined, affects the within-session reliability of RQA in an elderly population.Within-session reliability of RQA of CoP motion in the anterior-posterior and mediolateral directions was assessed in 267 individuals (>65 years old) when standing on firm or foam surface with eyes open or closed for each of two recurrence threshold methods. One threshold method sets the recurrence threshold level such that the recurrence rate is fixed to 5%, the other method sets the recurrence threshold based on 27% of the mean distance between all points from which recurrences are quantified. Reliability across four 30-s balance trials within each of four balance conditions (firm vs. foam, eyes open vs. closed) was determined using intra-class correlation, standard error of measurement and minimal detectable change.ICCs were better, the standard error of measurement and minimal detectable change were smaller when the recurrence threshold was set to 5% using the fixed recurrence threshold. Fixing recurrence rate improves the within session reliability of RQA and could increase sensitivity to identify fall risk.