Longitudinal study of free running exercise challenge: reproducibility.

The reproducibility of free running exercise challenge has been examined in an unselected population of 8-10 year olds. Using a standardised protocol, monthly exercise tests were performed on 143 children over one year. A positive test was defined using both a 15% and 20% fall in peak expiratory flow after exercise. The mean (95% confidence interval, CI) population frequency for a positive test at 15% fall was 14.9% (6.5 to 23.3) and coefficient of variation 24.6%. For a 20% fall, the mean (95% CI) population frequency was 7.9% (2.9 to 12.9) and coefficient of variation 27.8%. Seventy two (50.3%) of the children gave at least one positive response at 15% fall. Exercise testing is not reproducible in the community setting and should not be used as a screening test. Exercise data from epidemiological studies of asthma should be interpreted with caution.