Balance impairment and falls are not uncommon in later life. One in three individuals aged 65 and over experiences an accidental fall every year, with head injuries and hip fractures among the most severe consequences. Accordingly, several methods and techniques for assessing balance and fall risk have been developed. Posturography is probably the most common technique and entails the measurement of the body’s centre of mass (CoM) or the centre of pressure (CoP) displacements during standing.

An issue arises when researchers propose and compare new methods of posturography data analysis from different subjects across centres. These data are usually collected following different protocols. Moreover, there are also differences in the algorithms to process and characterise the data. This heterogeneity sometimes generates conflicting findings, most likely produced by a sizeable between-study variability and a low statistical power (i.e. small sample size). As a result, there is still a lack of consensus on the best methods to analyse posturography data in order to extract meaningful information about the subject's balance and fall risk.

Fortunately, in recent years some datasets of posturography evaluations have been made publicly available. In particular, researchers from the Biomechanics and Motor Control Laboratory at the Federal University of ABC in Brazil made available two datasets containing posturography data from cohorts including young and older adults, as well as sociodemographic and health characteristics. One dataset contains CoP data (doi:10.6084/m9.figshare.3394432.v2); the other one contains both CoP and CoM data (doi:10.6084/m9.figshare.4525082.v1). By giving different research groups access to the same data, they are providing the community with a potential normative reference for the comparison of different methods for data analysis and characterisation.

In this spirit, our group used the first dataset to compare the ability of linear and nonlinear descriptors of system dynamics to reveal differences in balance control between groups with different risk of falling. While young adults and older adults showed significant differences in linear measures, older adults with and without fall history over the past year showed similar CoP sways in terms of amplitude, variability and velocity of displacement (see Figure). In contrast, fallers showed more irregular CoP sways than non-fallers (i.e. higher entropy values), suggesting that nonlinear measures are more useful than linear measures to discriminate between these groups. These results were recently published in the Journal of NeuroEngineering and Rehabilitation, which can be accessed for free (https://rdcu.be/bdihg).

Finally, we invite other groups to join us in the use of public datasets for testing and comparing alternative methods for CoP and CoM data analysis and characterisation. In so doing they would be helping to reach the much-needed consensus for posturography-based balance and fall risk assessment.