Comparability of nonlinear measures of heart rate variability between long- and short-term electrocardiographic recordings

Abstract

N nonlinear measurements of heart rate variability (HRV) may reveal subtle changes in heart rate dynamics and the data support their superiority over conventional methods of HRV in predicting adverse cardiac events.1,2 Recent data on postinfarction patients with depressed left ventricular function suggest that the reduced short-term scaling exponent (a1) is a more powerful predictor of mortality than the traditional measurements of HRV.1,2 The studies evaluating the value of these methods in risk assessment have usually been done with 24-hour electrocardiographic (ECG) recordings.1,2 It would be important to know if similar information could be obtained from short-term ECG recordings. The comparability of conventional HRV measures derived from longversus short-term ECG recordings is quite good.3,4 However, the data on the comparability of the newer nonlinear measures of HRV obtained from longversus shortterm ECG recordings are very limited. The present study compares results of nonlinear HRV analysis performed in longversus short-term ECG recordings in healthy subjects and in postinfarction patients and determines which nonlinear dynamic parameters obtained from short-term recordings reliably reflect information provided by long-term recordings. • • • The study population consisted of 30 healthy subjects and 30 postinfarction patients. The healthy subjects (age 33 6 16 years [mean 6 SD], 15 men and 15 women, body mass index 24 6 3 kg/m [mean 6 SD], 9 smokers) had normal 12-lead ECG findings, did not have any evidence of disease or other health problems, did not have symptom, and were not on medication. The clinical characteristics of postinfarction patients are shown in Table 1. ECG data were recorded using the Burdick Holter recorder (Spacelab-Burdick, Milton, Wisconsin) allowing for 20 minutes of high-resolution research electrocardiography (1,000 Hz) followed by 24 hours of continuous ECG recording (200 Hz). ECG recordings were performed during the hospital stay for the postinfarction patients. However, for 24-hour and 20minute ECG recordings, the resolution of the tachogram used for HRV analysis was 5 ms. During the 20-minute recordings, the study subjects were in supine resting position. ECG data were transferred to a microcomputer for processing and HRV analysis. Premature beats and noise were excluded automatically and manually. HRV was analyzed by a software package with methods that have been previously described.5–8 Because of differences in autonomic tone under resting and ambulatory conditions, the 24-hour HRV data were compared with 2 different types of 10-minute HRV data: (1) 10-minute data obtained from the beginning of the 20-minute resting state recordings, and (2) 10-minute data obtained from the beginning of the 24-hour ambulatory data. All the time-domain and nonlinear measures of HRV were analyzed in 600-second blocks and the average values were used for comparisons. SD of all normal-to-normal RR intervals (SDNN), the square root of the mean squared differences of successive normal-to-normal RR intervals (RMSSD), and the proportion of interval differences of successive normal-to-normal RR intervals .50 ms (pNN50) were calculated as standard time-domain measures of HRV. The ratio of the low-frequency (LF, 0.04 to 0.15 Hz) component to the high-frequency (HF, 0.15 to 0.40 Hz) component of spectral power (LF/HF ratio) obtained by using the fast Fourier transformation method was determined as a frequency-domain measure of HRV. The analysis was done in 512-beat blocks. In the present study, 2 nonlinear HRV parameters From the Cardiology Unit, Department of Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York; and the Division of Cardiology, Department of Medicine, University of Oulu, Oulu, Finland. This study was supported in part by grants from the Academy of Finland, the Paavo Nurmi Foundation and the Maritza and Reino Salonen Foundation, Helsinki, Finland. Dr. Perkiomaki’s address is: Cardiology Unit, Department of Medicine, University of Rochester Medical Center, 601 Elmwood Avenue, Box 653, Rochester, New York 14642. E-mail: heartjsp@heart.rochester.edu. Manuscript received September 14, 2000; revised manuscript received and accepted October 26, 2000. TABLE 1 Clinical Characteristics of the Postinfarction Patients