The Effects of Sample Duration on the Reliability of Single Leg Balance Testing

Document Type

Article

Publication Date

2015

Publication Title

Medicine & Science in Sports & Exercise Supplemental

DOI

10.1249/01.mss.0000477000.92384.5f

ISSN

1530-0315

Abstract

Single leg balance testing (SLBT) is a commonly used tool in sports medicine; however, there has been no consensus on trial length needed to obtain reliable measures.

PURPOSE: To determine minimal trial length needed to obtain high relative and absolute reliability for SLBT.

METHODS: Three phases of SLBT were performed with the Biodex SD (BSD, Biodex, Inc., Shirley, NY) using dominant (D) and non-dominant (ND) limbs on stable (S) and unstable (US, BSD level 6) surfaces in a random order. Subjects (age 18-35yrs) were instructed to remain stable in a standardized testing position, with added instruction to keep the platform level during US surface testing. Each phase consisted of one practice trial and three test trials with 30s rest between trials. Phase 1 included healthy men (n = 5) and women (n = 5) maintaining balance up to 2 minutes for each limb/surface condition. Based on 85% of subjects maintaining balance for 45s without a compensatory event, Phase 2 involved different men (n = 15) and women (n = 15) performing 45s trials under each limb/surface condition. Based on reliability results of the overall stability index (OSI) for 6 cumulative time intervals (5s, 10s, 15s, 20s, 25s, 30s), trial length was established for Phase 3. Phase 3, designed to verify Phase 2 results, involved computing reliability across three trials by a different set of healthy men (n = 15) and women (n = 15). Due to heteroscedasticity, intraclass correlation coefficients (ICC, 2,3) and coefficient of variation (CV) were calculated on log transformed OSI for Phase 2 and 3 analyses.

RESULTS: Moderate to high ICC were revealed across Phase 2 times for S (D: 0.872- 0.911; ND: 0.895-0.941) and US (D: 0.798-0.902; ND: 0.736-0.865) surfaces. Although the 95% confidence intervals overlapped, ICC appeared to plateau at 15s suggesting minimal trial length needed. Phase 3 ICC results confirmed 15s trials had high relative reliability for S (D: 0.851; ND: 0.915) and US (D: 0.908; ND: 0.884) surfaces. High absolute reliability was supported through the CV for S (D: 23.0%; ND: 14.0%) and US (D: 13.8%; ND: 17.6%) surfaces.

CONCLUSION: When performing SLBT, 15s is the minimal trial length needed to obtain high relative and absolute reliability over three trials using D/ND limbs on S/US surfaces. Further analysis is required to determine reliability for single trials.

Comments

© 2015 American College of Sports Medicine

Link to Full Text

Share

COinS