By Dr Milton Maltz MD MPhil, London
Introduction
Repeated exercise testing is often employed for the assessment of the effects of anti-anginal agents and for the follow up of patients with angina pectoris. However, various external factors such as environmental temperature (Epstein et al 1971), post prandial state (Simonson and Keys 1960), emotional stress (Robinson 1968) and cigarette smoking (Aronow and Kaplan 1968) have been shown to affect exercise response. Although there appears to be no significant variation in the results of exercise tests performed at the same time on different days (Dagenais et al 1971), recent reports suggest the possibility of diurnal fluctuation of the maximum heart rate and degree of ST segment depression in response to exercise (Joy et al 1982, Reilly et at 1983).
The present chapter described the reproducibility of exercise testing in 20 patients with stable angina and proven coronary artery disease. Ten of these patients underwent control treadmill exercise tests on different days within one week, the other 10 patients underwent 5 exercise treadmill tests on the same day (before placebo and at 1, 3, 6, and 9 hours after placebo).
Statistical Methods
For measurements recorded on different days, between day variation was assessed. The Wilcoxon test was used to assess any difference between coefficients of variation for double product and exercise duration. For each variable of interest, analysis of variance was used to compare measurements taken at different times of the day within the same patient to their pre-place value. Results were taken as significant at the 5% level.
Results
Variability during the day
The results of heart rate and blood pressure at rest and maxiam exercise with double product are summarised in table 4. At maximal exercise the double product, used as an index of oxygen consumption, did not differ significantly at different times of the day as compares to the pre-placebo (baseline) exercise test. Resting heart rate and resting blood pressure did also not differ during the day from their baseline values.
The time intervals to reach specific events of the exercise treadmill test are summarised in table 5, figures 9, 10 and 11. With the same exercise protocol (chapter 2), exercise duration, the time to the onset of moderate angina, the time to noser of 1mm ST segment depression, maximum ST depression and time for the ST segment to return to normal were similar at different times of the day. There were a trend of a better exercise performance at the midday test, although this did not reach statistical significance. During exercise, time to onset of maximum ST segment depression showed the greatest variability. However, the range of absolute valuers was widest for the time to onset of 1mm ST depression.
Variability within an individual
The standard deviation of the double product at maximal exercise expressed as a percentage of the mean double product)coefficient of variation) was calculated for each patient. The median coefficient of variation for the double product was 6.3% (range 3.4% to 12.6%). The median coefficient of variation of exercise duration was 9.9% (range 4.6 to 28%). There was a statistically significant difference between these two coefficient (p<0.03), (table 6). Tables 1A and 2A (Appendix A) summarise the percentage of changes of resting double product and exercise duration respectively in each of the ten patients. In absolute terms, up to 23% in variability from baseline values my be recorded in any one subject using identical exercise protocols. Even greater variability was recorded in exercise duration, with one subject varying by up to 40%.
Variability between weeks
Analysis of variance was performed for the 5 successive control tests to assess if there was a change from the first test due to training or familiarisation effects during serial exercise testing. There was no evidence of a training effect during testing (table 7).
Discussion
Exercise tolerance tests are often used to evaluate the effectiveness of anti-anginal drugs, but interpretation of results may be difficult because of possible day-to-day and test-to-test variation and the possible effect of training on the results of exercise tests repeated at short intervals (Hame et al 1966, MacAlpin et al 1966, Robinson et al 1968). However, it has been shown that angina occurs at reproducible rate-pressure products despite changes in the external work performance (Debri et al 1967, Gobel et al 1978, Robinson et at 1967). However, they also observed that the time to reach 1mm ST segment depression exhibited more variability with the time of the day.
In the present study, greatest diurnal variability was shown by the time to maximum ST segment depression, although the differences did not reach statistical significance. Both the rate pressure product and exercise duration at maximal exercise similarly showed some but not significant diurnal variation with a trend to a midday peak.
Furthermore, when comparing the coefficient of variation over the day, both double product and exercise duration exhibited wide inter-patient variability (table 6). Overall, the above suggests that the duration of action of anti-anginal medications can be assessed by repeated exercise testing on the same day but a true “training effect” may by obscured, by a small study sample size or wide individual variation. This has been noted by Dagenais et al 1969. Hamer et al 1966 and MacAlpin et al 1966.
Other authors (MacAlpine et at 1966, Crea et al 1986, Sugishita et al 1986) have suggested that alterations in vascular time are among potential mechanisms for diurnal variations.
These factors in addition to the above discussed points should be borne in mind when assessing any individual patient’s response to an anti-anginal medication.