Repeatability of measurements and sources of variability in tests of cardiovascular autonomic function.

OBJECTIVE To determine the repeatability and sources of variability of clinical tests of cardiovascular autonomic function. DESIGN The commonly used electrocardiographic related tests of autonomic function were studied. Two repeat measurements of all tests were made on all subjects on four separate days over a four week period. SUBJECTS Ten normal subjects with no known autonomic dysfunction were investigated. MAIN OUTCOME MEASURES These were deep breathing (subject seated and supine), Valsalva manoeuvre, standing up from lying position, and normal relaxed breathing (subject supine). During the tests the electrocardiogram and respiratory pattern were recorded by computer. Beat to beat RR intervals were measured automatically from the electrocardiogram, and from these the results of the tests were calculated. RESULTS Variance analysis showed significant between subject variability for all tests (p < 0.005), but some tests showed a much smaller relative within subject variability than others. Average repeatability data (within subject SD) for each test were calculated, and included deep breathing sitting (max-min) RR (46 ms), Valsalva ratio (0.17), and lying to standing RR ratio (0.11). These compare with between subject SDs of 65 ms, 0.38, and 0.13 respectively, at mean values of 305 ms, 1.92, and 1.15 respectively. The data highlighted one subject with the poorest repeatability, whose electrocardiogram turned out on closer inspection to be under atrial rather than sinus control at times. Poor repeatability in the other subjects was related to variability in the respiratory pattern, and in the deep breathing test, repeat variability was significantly correlated (r = 0.79) with variability in the respiratory amplitude (p < 0.05). CONCLUSIONS Repeatability data should be available to each laboratory carrying out autonomic function tests. The data provided in this study could be used as a baseline. Poor repeatability highlights the need to re-examine the test procedures, or the test data from specific subjects. Variability of respiratory pattern is associated with poor repeatability, and so careful instructions on respiration should be given to each subject before the tests.