Interleg Coordination in Quiet Standing: Influence of Age and Visual Environment on Noise and Stability

ABSTRACT

The authors reexamined reported effects of age, illumination, and stationary visible structure on the net center of pressure (COP) derived from dual, side-by-side force plates (J. Kinsella-Shaw, S. Harrison, C. Colon-Semenza, & M. Turvey, 2006) from the perspective of axial postural control. They questioned how left and right COP _x (t), COP _y (t), and vertically oriented ground reactive force, GRF _z (t), coordinated during quiet standing. The Cross- recurrence Quantification (CRQ) revealed that coordination was primarily between fluctuations of similar direction, with coordination of left and right COP _y (t) (anteroposterior fluctuations) dominant. CRQ also revealed that (a) illumination and structure affected the interlimb dynamics of older (M age = 72.2 ± 4.90 years) participants more than their younger (M age = 22.8 ± 0.83 years) counterparts, and (b) older participants exhibited greater interlimb entrainment (dynamical stability) in the presence of greater interlimb noise.

Keywords:

• aging
• illumination
• interlimb coordination
• nonlinear dynamics
• postural control

ACKNOWLEDGMENTS

A grant (Collaboratory for Rehabilitation Research) awarded by the Provost's Office of the University of Connecticut supported this research.

Notes

1. In simplest terms, the %Cross RECUR measure provides an estimate of the prevalence of noise in the time series. Consider that what the measure does is evaluate the number of times a system revisits the same states during a given interval. The fewer returns to previously visited states that are observed, the more states the system must have occupied over the interval of observation and the more random or noisy is its behavior. The recurrence-derived measure on Shannon entropy provides an allied index on this version of noise. In the extreme case of white noise, all possible states would be represented in the interval evaluated with an equal probability of occurrence at any time. In the context of postural control and other voluntary actions, noise of this kind can be conceived of as any trajectories observed that are not congruent with those required by the behavioral goal. In physical systems, the origins of noise are thermal. For living systems the origins of this noise ultimately resides in the metabolic processes that liberate biochemical energy sufficient to maintain viability. As these underlying processes are cyclical and operating at multiple time scales, they are correlated at multiple time scales and display fractality. In the context of movement trajectory generation, this plays out in the motor system in the interaction of baseline physiological tremor, the tonic contractile states required for dealing with gravity, and the phasic contractile states required for torque generation around joint axes that result in voluntary movement. Additionally, to the extent that information can change the scaling relations across these levels of the motor system, changes in the sensitivity of sensory detection (e.g., visual, proprioceptive, vestibular) can serve to increase the noise associated with movement generation. Shaw and Kinsella-Shaw (1988) provided formal grounds for this in the context of an analysis of goal-directed behaviors. Briefly, some degree of movement trajectory overshoot and undershoot is always observable as the different levels of the sensory and motor systems differ in their resolving power or availability. For these reasons, the well-documented changes in sensory detection capacity that accompany aging should be accompanied by corresponding changes in noise, as evaluated under cross- recurrence quantification analysis.

2. In the present context, the cross-recurrence measures are most usefully (and conservatively) viewed as providing information about the recorded movements’ dynamics, originating in a three-way interaction among the task constraints, the mechanical properties of the body, and the contribution of the neural substrate as it supports information detection and action. RQA methods have been applied to postural fluctuation data collected from individuals with Parkinson's disease, diabetic peripheral neuropathy, and clinically significant age-related changes in sensory and motor capacities (Bonnet, Carello, & Turvey, 2009; Kinsella-Shaw et al., 2006). In the broader neurological domain, RQA has also been applied to data collected from normally developing infants, as well as individuals with cortical disease, cardiac arrhythmia, and epilepsy (Aßmann, Romano, Thiel, & Niemitz, 2007; Babloyantz, 1991; Bianciardi et al., 2007; Censi et al., 2000; Harrison, Frei, & Osorio, 2008). These studies suggest that recurrence quantification measures are sensitive to changes in the functional connectivity of the nervous system. Furthermore, studies of Parkinsonian postural control and work with multiple electrode array chips (designed to provide simple simulations of cerebellar circuits) suggest that RQA measures may be particularly responsive to changes in the modulatory contributions of subcortical structures (i.e., basal ganglia and cerebellum) to postural control (Gour et al., 2007; Novellino, & Zaldívar, 2010).