Assessment of heart rate variability in hypertrophic cardiomyopathy. Association with clinical and prognostic features

Abstract

Background: Altered vascular responses during exercise and disturbed responses to autonomic function testing have been documented in hypertrophic cardiomyopathy (HCM) and are associated with markers of an adverse prognosis. Reduced heart rate variability (HRV) and baroreflex sensitivity are predictors of increased risk of sudden death after myocardial infarction, but the value of these parameters in HCM is unknown.

Methods and results: To determine the clinical significance of HRV and its relation to markers of electrical and hemodynamic instability in HCM, the 24-hour Holter recordings of 104 patients in sinus rhythm and off medication were analyzed. Five nonspectral measures of HRV were computed. The frequency components of HRV were calculated by fast Fourier transformation of the RR time intervals; the areas under the low (0.04 to 0.15 Hz) and high (0.15 to 0.4 Hz) frequency portions of the spectrum were measured as indices of autonomic and specific vagal influences on HRV, respectively. Spectral and nonspectral measures were compared with clinical, echo/Doppler, and Holter variables. ANCOVA was performed to allow for the effect of age on differences between variables. Spectral and nonspectral measures of HRV were correlated (r > .65; P < or = .001), indicating that the different time-domain and frequency parameters reflected similar measures of HRV. Global measures of HRV including the standard deviation of the mean of RR intervals (SDRR) and the standard deviation of 5-minute mean RR intervals (SDANN) were increased in patients with an adverse family history of HCM (173 +/- 67 vs 131 +/- 38 milliseconds, P = .001, and 158 +/- 66 vs 116 +/- 36 milliseconds, P = .004, respectively). In patients with exertional chest pain, global nonspectral measures were reduced compared with asymptomatic patients (118 +/- 31 vs 152 +/- 53 milliseconds, P = .006, and 105 +/- 30 vs 136 +/- 52 milliseconds, P = .014, respectively). Specific vagal influences on HRV including the proportion of RR intervals more than 50 milliseconds different (PNN50) and the high frequency peak on spectral analysis were less in patients with supraventricular arrhythmias on Holter monitoring (7.2 +/- 8 vs 16 +/- 13%, P = .012, and 21 +/- 10 vs 28 +/- 13 milliseconds, P = .048, respectively). Similarly, both global and specific vagal measures of HRV were less in the 27 patients with nonsustained ventricular tachycardia on Holter (PNN50, 7.7 +/- 9 vs 15 +/- 13 milliseconds, P = .048, and high frequency component, 19 +/- 9 vs 28 +/- 13 milliseconds, P = .05. During follow-up, 10 patients, 9 of whom were aged less than 33 years, experienced catastrophic events; 6 were resuscitated from ventricular fibrillation and 4 died suddenly. Indices of HRV were similar in these 10 patients to indices in the 94 survivors.

Conclusions: Time-domain and spectral measures of HRV yield similar information about the specific autonomic influences on the heart. Global and specific vagal influences on HRV were reduced in patients with symptoms and arrhythmias and global HRV is increased in patients with an adverse family history of HCM, but these indices do not add to the predictive accuracy of established risk factors.