The spectral characteristics of postural sway behaviour

  • R. W. Soames
  • J. Atha

DOI: 10.1007/BF02334065

Cite this article as:
Soames, R.W. & Atha, J. Europ. J. Appl. Physiol. (1982) 49: 169. doi:10.1007/BF02334065


To determine whether recognisable profiles of sway behaviour are characteristically found among normal subjects the energy content of a contiguous series of bandwidths were computed from power density spectra of the sway behaviour of 29 young men and 29 young women. No significant differences were observed over time either in the energy content or in the profile shape of the power spectral density (PSD) records when these were subjected to a multivariate profile analysis of variance.

It was observed that the majority of energy in the power spectrum, viz. 90% (±4%) in the antero-posterior and 95% (±3%) in the lateral direction, was below 2.0 Hz. Three dominant frequencies of sway were identifiable in 93% of subjects. The modal frequencies in the antero-posterior direction were 0.30–0.45 Hz (primary), 0.60–0.75 Hz (secondary) and 1.05–1.20 Hz (tertiary), and in the lateral direction 0.30–0.45 Hz (primary), 0.45–0.60 Hz (secondary) and 0.75–0.90 Hz (tertiary). These three dominant frequencies were distributed over the full frequency range. An explanation of their origins was not sought but the suggestion is made that they could be driven by common dynamic physiological events.

The patterns of energy distribution were simplified and quantified by the calculation of a difference-ratio index. The broad dispersion of the values of this index precludes the suggestion that there could be any clustering of profiles.

It is concluded that the frequently postulated clustering of the sway patterns of young adults into recognisable and distinct profiles is without firm foundation. Instead it is suggested that sway behaviour is probably normally and randomly distributed about common respiratory, cardiac and perhaps also other dynamic physiological events.

Key words

Postural sway Steadiness Sway frequency Power spectral density 

Copyright information

© Springer-Verlag 1982

Authors and Affiliations

  • R. W. Soames
    • 2
  • J. Atha
    • 2
  1. 1.Dept. of AnatomyUniversity of London King's CollegeLondonEngland
  2. 2.Dept. of Human SciencesLoughborough University of TechnologyLoughboroughEngland

Personalised recommendations