Harbourne, R T.; DeJong, S L.; Kurz, M J.

Pediatric Physical Therapy: Spring 2005 - Volume 17 - Issue 1 - p 70
Section Information: Abstracts of Platform and Poster Presentations for the 2005 Combined Sections Meeting: Poster Presentations

Munroe-Meyer, Physical Therapy, Univ. of Nebraska Medical Center, Omaha, NE, USA (Harbourne, DeJong)

University of Nebraska at Omaha, Omaha, NE, USA (Kurz)

PURPOSE/HYPOTHESIS: It is well known that chaos is a central feature of human locomotion. A chaotic gait pattern means that the lower extremity kinematics fluctuate from one step to the next with a deterministic pattern. Although such fluctuations have provided insight on the stability and health of the neuromuscular system, less is known about the emergence of chaotic gait patterns in typically developing infants. This investigation represents our initial efforts to explore the development of a chaotic gait pattern in a healthy infant. The purpose was to further explore variability in the walking cycle as independent walking emerged.

NUMBER OF SUBJECTS: Longitudinal data collection from one normal infant from 11 months (pre-walking) to 14 months, at weekly intervals, for a total of 15 sessions.

MATERIALS/METHODS: 2-D motion analysis using video and the Peak Performance Technologies motion capture system. The infant was lightly supported at the arms by his parent during treadmill walking at 0.4 mph. Reflective markers placed at the greater trochanter, knee and ankle defined the variables of knee angle and leg pendulum. Time series of these variables for each session were selected with at least 30 consecutive footfalls. The Lyapunov Exponent (LyE) was calculated to determine the chaotic structure of the respective time series. The LyE for a periodic signal (e.g. sine wave) is zero, and the LyE for random noise is positive (+0.469). A chaotic gait pattern lies between the two extremes. Shifts of the LyE towards randomness may indicate instability, while shifts towards periodicity are associated with stability or even rigidity.

RESULTS: For both the knee angle and leg pendulum, the LyE shifted towards randomness over time, with greater periodicity (0.201 for knee angle and 0.104 for leg pendulum) at the earliest, pre-walking sessions. LyE shifted to 0.277 for knee angle and 0.176 for leg pendulum by 14 months, indicating a more random pattern after the infant had learned to independently walk.

CONCLUSIONS: The pre-walking sessions may have had greater periodicity in the movement pattern due to the infant’s reliance on the driving motion of the treadmill to produce a viable walking pattern, and the infant’s dependence on that movement because independent walking had not yet emerged. Over time, the child’s walking pattern shifted towards randomness. This may reveal the child’s exploration of his body’s degrees of freedom and options for control using a variety of strategies. Our findings support other research in dynamic systems indicating that higher instability and chaotic structure in movement is characteristic of normal emergent behavior.

CLINICAL RELEVANCE: With pediatric intervention utilizing treadmill training to improve walking, it is necessary to understand the nature of variability in emergent walking behavior in the normal infant. The importance of increasing variability as a part of that emergent behavior should be incorporated into treatment and the process of evaluation of the results of intervention.

© 2005 Lippincott Williams & Wilkins, Inc.