Do regression-based computer algorithms for determining the ventilatory threshold agree?

Panteleimon Ekkekakis, Erik Lind, Eric Hall, Steven Petruzzello

Research output: Contribution to journalArticlepeer-review

Abstract

The determination of the ventilatory threshold has been a persistent problem in research and clinical practice. Several computerized methods have been developed to overcome the subjectivity of visual methods but it remains unclear whether different computerized methods yield similar results. The purpose of this study was to compare nine regression-based computerized methods for the determination of the ventilatory threshold. Two samples of young and healthy volunteers (n = 30 each) participated in incremental treadmill protocols to volitional fatigue. The ventilatory data were averaged in 20-s segments and analysed with a computer program. Significant variance among methods was found in both samples (Sample 1: F = 11.50; Sample 2: F = 11.70, P < 0.001 for both). The estimates of the ventilatory threshold ranged from 2.47 litres · min-1 (71% VO2max) to 3.13 litres · min-1 (90% VO2max) in Sample 1 and from 2.37 litres · min-1 (67% VO2max) to 3.03 litres · min-1 (83% VO2max) in Sample 2. The substantial differences between methods challenge the practice of relying on any single computerized method. A standardized protocol, likely based on a combination of methods, might be necessary to increase the methodological consistency in both research and clinical practice.

Original languageEnglish (US)
Pages (from-to)967-976
Number of pages10
JournalJournal of Sports Sciences
Volume26
Issue number9
DOIs
StatePublished - Jul 2008

Keywords

  • Anaerobic threshold
  • Computer algorithms
  • Gas exchange threshold
  • Limits of agreement

ASJC Scopus subject areas

  • Orthopedics and Sports Medicine
  • Physical Therapy, Sports Therapy and Rehabilitation

Fingerprint

Dive into the research topics of 'Do regression-based computer algorithms for determining the ventilatory threshold agree?'. Together they form a unique fingerprint.

Cite this