Use of recovery VO2 to predict running economy

The purpose of this study was to determine whether running economy (RE) could be predicted accurately using recovery VO2 values. Twelve runners (means VO2max = 61.9, SD 4.9 ml.kg-1.min-1) completed three treadmill RE sessions over a 2-week period. During each session, subjects performed three 6-min runs at 69%, 78%, and 87% VO2max. RE was calculated from a single 2-min gas collection during the last 2 min of running. Immediately following each run, recovery VO2 data obtained during randomly assigned 15-s, 20-s, or 25-s gas collections were used to predict exercise VO2. Correlations and mean absolute percentage variation (% VAR) between actual and predicted VO2 at each relative intensity and recovery period are reported. Although the relationship between actual and predicted VO2 was significant and more pronounced at higher exercise intensities, the overall magnitude of the association was low to moderate (r range = 0.50-0.81). The range of % VAR between actual and predicted aerobic demands also obscured marked underprediction (-6.5% to -12.5%) and overprediction (+10.1% to +17.4%) of actual VO2 in some subjects. These data suggest that 15-, 20-, and 25-s recovery VO2 values do not correlate strongly with steady-state VO2, nor do they adequately account for variation in individual economy profiles.