LABORATORY MEASURES ARE NOT VERY SENSITIVE TO REAL PERFORMANCE CHANGES
Garner, A. S., Martin, D. T., Gulbin, J., Doney, G. E., Jenkins, D. G., & Hahn, A. G. (2003). Laboratory and velodrome sprint cycling power in female cyclists following 6 weeks of training. Medicine and Science in Sports and Exercise, 35(5), Supplement abstract 1872.
Female cyclists (N = 21) completed six weeks of training in a velodrome following one week of familiarization. Before and after training, 30-s air-braked cycle ergometer tests (S-LAB) and a maximal 500 m time trial (S-VELO) were performed."The relationship between change in speed and change in power . . . during S-VELO was strong . . . However, the relationship between change in S-VELO speed and S-LAB power was moderate . . . and small between change in power . . . for S-LAB and S-VELO.
The small to moderate relationship between laboratory sprint data and changes in 500 m TT [time trial] performance suggest that other factors such as technique influence adaptation to training. Fitness traits that predict performance may not always predict changes in performance." (p. 337).
Implication. Laboratory measures are not very sensitive to real-life performance changes in cyclists.