Biomechanical approach to quantifying anticipatory postural adjustments in the elderly.

The paper outlines a biomechanical approach to quantifying anticipatory postural adjustments in the elderly. The measurement problems that occur in applying the biomechanical approach to elderly subjects are described and the 'signal-to-noise' properties of three candidate measures are compared, using data from volitional unilateral arm-raise tests performed on 100 elderly subjects. The results suggest that changes in vertical ground reaction force provide the greatest potential for accurate measurement of anticipatory adjustments, in comparison with changes in horizontal force or centre-of-pressure displacement. By normalising the anticipatory change in vertical ground reaction force with respect to the vertical perturbation force induced by the arm motion, a measure of relative anticipatory response is derived. The use of this measure, as well as its limitations, are demonstrated by analysing its relationship to actual falling risk, monitored prospectively in the elderly subject population. The findings showed evidence of larger relative anticipatory adjustments in the subjects who experienced recurrent falls, and it is suggested that these responses may be indicative of disordered motor programming. However, to detect these differences, it was necessary to average responses over multiple trials and to exclude trials with very small arm acceleration or very large baseline 'noise' (associated with ongoing postural sway). The need to screen and exclude data would seem to limit the practical utility of this approach in testing elderly populations.