Short communication Local dynamic stability of the lifting kinematic chain Ryan B. Graham a, *, Patrick A. Costigan a,b , Erin M. Sadler a , Joan M. Stevenson a,b a School of Kinesiology and Health Studies, Queen’s University, 28 Division Street, Kingston, Ontario, Canada, K7L 3N6 b School of Rehabilitation Therapy, Queen’s University, 31 George Street, Kingston, Ontario, Canada, K7L 3N6 1. Introduction During goal-directed tasks, such as low-lying object lifting, the human motor system regulates the body’s centre of mass (CoM) and trunk position to maintain equilibrium and prevent falling [1]. During lifting any voluntary trunk movement will perturb that equilibrium [2]. This perturbation will trigger an automatic postural response [3] requiring coordinated control of the spine, pelvis, and lower limb [4]. Accurate neuromuscular control allows these involuntary postural responses to be smoothly incorporated into movement repertoires to ensure precise, harmonious motion [2,3]. When the perturbations are increased by such things as external loads, the corresponding postural adjustments become more complex, involving more joints and muscles, increased muscle activation, and different activation strategies such as increased co-contraction [2]. While walking, the central nervous system (CNS) prioritizes the stability of the superior over the inferior segments [5]; however, what is prioritized for lifting is unknown. Therefore, we examined foot, shank, thigh, pelvis, lower back, and upper back segment stability during repetitive symmetric lifting from floor to waist height. Because active neuromuscular control of trunk motion is as crucial in lifting as it is in walking, we hypothesized that stability would increase when moving up the kinematic chain [5]. 2. Methods Thirty healthy volunteers (15M, 15F) participated after providing informed consent approved by the University Research Ethics Board. Participants’ mean age, height, and weight were 24.2(2.9) years, 184.9(7.6) cm and 85.4(10.7) kg for males; and 23.0(2.6) years, 170.1(6.1) cm and 66.3(11.7) kg for females. The lifting protocol is described elsewhere in detail [6–8]. Subjects performed 30 continuous, symmetrical, freestyle box lifts with their feet stationary (10/min) from a target on the floor to a target on a table top, which was set at 50% of their standing height. The load lifted was equivalent to 10% of each participant’s maximum back strength; established previously [6–8]. The average load lifted was 6.93(0.72) kg for males, and 4.8(0.65) kg for females. Segment kinematics were collected using an Optotrak 3020 System (Northern Digital Inc., Waterloo, ON, Canada) at 100 Hz. Six infrared emitting diode (IRED) marker triads were located laterally on each subject’s right foot, shank, and thigh, as well as on custom-made fins projecting from S 1 , T 12 , and C 7 . Custom Matlab software (The MathWorks, Natick, MA, USA) was used for all data processing [6,7]. Right-handed technical coordinate systems were defined from each triad to track the linear and rotational motions of each segment. Linear motions were defined using the origin of each coordinate system, whereas 3D rotations were defined using a tilt-obliquity-rotation convention relative to the laboratory global coordinate system [5]. Both linear and angular data were then filtered using a 10 Hz low-pass second-order digital Butterworth filter [5–7,9], and converted to velocities to reduce non-stationarities in the displacement data [5,10]. The last 25 lifts were extracted to ensure a steady-state movement pattern [9], and the data were time normalized to a constant number of data points (15,000) as the number of samples can affect stability analyses [11]. For each segment an attractor was reconstructed by defining a 12-dimensional state space using linear and angular velocities ½ ˙ x˙ y˙ z ˙ u ˙ ’ ˙ c and their time-delayed copies [5] (Eq. (1)): YðtÞ ¼ ½ ˙ xðtÞ ˙ yðtÞ ˙ zðtÞ ˙ uðtÞ ˙ ’ðtÞ ˙ cðtÞ ˙ xðt  t 1 Þ ˙ yðt  t 2 Þ ˙ zðt  t 3 Þ ˙ uðt  t 4 Þ ˙ ’ðt  t 5 Þ ˙ cðt  t 6 Þ (1) Time delays were calculated for each state variable using the first minimum of the average mutual information function [12,13]. Short- and long-term maximum Gait & Posture 34 (2011) 561–563 A R T I C L E I N F O Article history: Received 1 November 2010 Received in revised form 10 May 2011 Accepted 29 June 2011 Keywords: Lifting Stability Neuromuscular control Lyapunov exponents Kinematic chain A B S T R A C T While a stable trunk and centre of mass (CoM) trajectory are required during lifting, it is unclear how stability is controlled. Thirty healthy participants (15M, 15F) performed repetitive, symmetric lifting at 10 cycles per minute for 3 min with a load-in-hands equivalent to 10% of their maximum back strength. Short- and long-term maximum finite-time Lyapunov exponents (l max-s and l max-l ), describing responses to small (local) perturbations, estimated the local dynamic stability of the foot, shank, thigh, pelvis, lower back, and upper back segments. Instability (l max-s ) significantly increased when moving up the kinematic chain (p < 0.001). Therefore, to maintain trunk equilibrium and accurately regulate CoM trajectory during lifting, stability of the distal (fixed) lower limb segments is prioritized. This is contrary to previous results observed during gait, indicating that trunk control via kinematic chain stability is accomplished differently for walking and lifting. ß 2011 Elsevier B.V. All rights reserved. * Corresponding author. Tel.: +1 613 533 6000x79019; fax: +1 613 533 2009. E-mail address: ryan.graham@queensu.ca (R.B. Graham). Contents lists available at ScienceDirect Gait & Posture jo u rn al h om ep age: ww w.els evier.c o m/lo c ate/g aitp os t 0966-6362/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2011.06.022