Article Text

PDF
12 Movement pattern clustering of patients with self-reported ankle instability during a jump task
  1. JT Hopkins1,
  2. H Kim1,
  3. SJ Son1,
  4. S Reese2,
  5. R Roundy2,
  6. MK Seeley1
  1. 1Human Performance Research Center, Brigham Young University, Provo, UT, USA
  2. 2Department of Statistics, Brigham Young University, Provo, UT, USA

Abstract

Background Self-report instruments classify ankle instability (AI) by self-perceived symptoms and movement deficits however, movement neuromechanics in this patient population are quite variable. For this reason it is difficult to identify specific neuromechanical alterations that might lead to poor movement strategies and perpetuate AI.

Objective To identify multiple clusters of lower extremity neuromechanics, representing movement patterns in patients with AI.

Design Descriptive laboratory study.

Setting Research laboratory.

Patients 100 subjects (22 ± 2 yrs) with a history of ankle sprains (4.4 ± 3.2), scored below 90% (83 ± 9) on the FAAM ADL, below 75% (62 ± 13) on the FAAM Sport, reported at least 2 “yes” responses (4 ± 1) on the MAII, and had no sprain in the previous 3 months.

Interventions High-speed video (250 Hz) and force plate (2500 Hz) data were used to estimate sagittal and frontal plane ankle, knee, and hip angles and joint torques during 5 trials of a max vertical jump, onto a force plate, immediately followed by a lateral jump. EMG data (2000 Hz) were collected from lower extremity muscles.

Main outcome measurements A curve or “function” was generated for each trial collected during the stance phase of the jump task. Relevant DVs were grouped according to their relationship of a movement pattern. Bayesian statistical modelling was used to create clusters from the functions. Functions were clustered using the multiple landmarks that characterised the grouped function. Cross-classification matrices were used to measure cluster dependences across the grouped variables.

Results Distinct clusters were identified among the grouped variables. Summary functions/curves represent each distinct cluster. The maximum proportion of subjects sharing similar clusters from at least two different grouped variable characteristics (joint and plane) ranged from 0.10 to 0.58.

Conclusions Multiple distinct joint torque functions were identified in a “homogenous” group of patients with self-reported AI, suggesting that multiple distinct neuromechanical alterations exist in an AI patient population. These data should be considered when using self-report instruments to identify AI, as these instruments may not be sensitive enough to provide evidence of a single movement strategy that places AI patients at risk. More data are needed to provide clinical identification techniques for each of the data clusters.

Statistics from Altmetric.com

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.