Test-retest repeatability of muscle sympathetic nerve activity: influence of data analysis and head-up tilt.

Total integrated muscle sympathetic nerve activity (MSNA) is composed of bursts that vary in both frequency and amplitude. Various quantifiable indices are currently used to characterize MSNA and its reflex-mediated responses. However, a comprehensive and systematic analysis on the test-retest repeatability of these measures has not been conducted. Therefore, the purpose of this study was to compare the consistency of supine and passive head-up tilt-mediated sympathetic nerve activity using different descriptors of MSNA and a statistical paradigm that included Model II ordinary least products (OLP) regression, Bland-Altman method of differences, and analysis of variance. MSNA (microneurography), stroke volume (SV, Doppler), and arterial blood pressure (ABP, Finapres) were measured during repeated supine and 60 degrees head-up tilt (HUT) conditions separated by a minimum of 3 weeks. MSNA was quantified using; burst frequency (and incidence), burst amplitudes (and total integrated activity) normalized to the largest absolute amplitude within each posture, and calculated percent changes (from supine) in absolute burst amplitude and total integrated activity. Most indices of MSNA showed excellent test-retest repeatability during both postures with neither fixed nor proportional bias. However, MSNA expressed as burst incidence demonstrated both fixed and proportional bias in the supine position, but not during HUT. In addition, HUT-induced percent changes in absolute burst amplitude and total activity displayed a fixed bias with greater increases during the second test (P<0.05). The hemodynamic variables associated with the reflex responses were quite similar between tests (i.e., no bias). It was concluded that, with the exception of burst incidence, the majority of MSNA indices provided reliable markers of sympathetic activity on repeated tests. However, care must be taken when using percent changes in MSNA that incorporate absolute amplitudes.