The ability of a submaximal cycle ergometer test to detect longitudinal changes in VO(2)max

BACKGROUND: The purpose of the present study was to examine the ability of a submaximal cycling test to detect longitudinal changes in maximal oxygen uptake (VO(2)max) and examine the conformity between changes in measured and estimated VO(2)max over a time span of 5–8 years. METHODS: A total of 35 participants (21 men and 14 women), aged 29 to 63 years, performed the Ekblom-Bak (EB) submaximal cycle test for estimation of VO(2)max and a maximal treadmill running test for direct measurement of VO(2)max. The baseline tests were conducted between 2009 and 2012, and the follow-up tests were completed 5 to 8 years later. Pearson’s coefficient of correlation (r) and paired sample t-test were used to analyse the association between change in measured and estimated VO(2)max. Random and systematic errors between the measured and estimated VO(2)max were evaluated using Bland-Altman plots. Repeated measures ANOVA were used to test differences between changes over time. RESULTS: There was no significant change in mean measured VO(2)max between baseline and follow-up (p = 0.91), however large individual variations were noted (− 0.78 to 0.61 L/min). The correlation between individual change in measured and estimated VO(2)max was r = 0.75 (p < 0.05), and the unstandardised B-coefficient from linear regression modelling was 0.88 (95% CI 0.61 to 1.15), i.e., for each litre of change in estimated VO(2)max, the measured value had changed 0.88 L. The correlation between baseline and follow-up errors (the difference between estimated-measured VO(2)max at each occasion) was r = 0.84 (p < 0.05). With regard to the testing procedure, repeated measures ANOVA revealed that there was no significant difference between the group who exercised at the same work rates at baseline and follow-up (n = 25), and those who required a change in work rate (n = 10). CONCLUSIONS: The EB test detected a change in VO(2)max with reasonably good precision over a time span of 5–8 years. Further studies are needed to evaluate if the test can be used in clinical populations and in subjects with different medications.