New method to measure and improve consistency of baroreflex sensitivity values

• Published in: Clinical autonomic research Vol. 20; no. 6; pp. 353 - 361
• Main Authors: Bernardi, Luciano, De Barbieri, Giacomo, Rosengård-Bärlund, Milla, Mäkinen, Ville-Petteri, Porta, Cesare, Groop, Per-Henrik
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 01.12.2010; Springer Nature B.V
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Algorithms; Autonomic Nervous System - physiology; Baroreflex - physiology; Blood Pressure - physiology; Cardiology; Child; Child, Preschool; Data Interpretation, Statistical; Diabetes; Endocrinology; Female; Gastroenterology; Heart Rate - physiology; Humans; Infant; Linear Models; Male; Medicine; Medicine & Public Health; Middle Aged; Neural Networks (Computer); Neurology; Ophthalmology; Posture - physiology; Reference Standards; Research Article; ROC Curve; Young Adult; Baroreceptors; Blood pressure; Autonomic nervous system; Neural networks; Heart rate variability
• ISSN: 0959-9851; 1619-1560; 1619-1560
• DOI: 10.1007/s10286-010-0079-1

Objective Baroreflex sensitivity (BRS) is an important prognostic index in cardiovascular diseases, however, its use is complicated by different methods difficult to compare and standardize, often providing conflicting results. We tested whether the simple ratio of RR interval to systolic blood pressure global variabilities (assessed by standard deviations) is a reliable measure of BRS, by measuring the agreement with six established methods. In addition, we tested whether high-pass filtering of data, by removing slow non-baroreflex-mediated fluctuations, could improve the agreement between different BRS methods. Methods In 1,409 subjects, we compared 6 established methods (derived by cross-spectral and sequence analysis) and the new method, supine and in response to tilting (1,175 subjects). Data were analyzed after linear detrending, high-pass filtering at 0.025 and 0.05 Hz. Results Although all seven methods showed a general agreement, the new method consistently showed the lowest distance from the median of the remaining methods (0.04 ± 0.06 ms/mmHg over 2,584 files, p  < 0.05 with respect to the second best method). High-pass filtering improved ( p  < 0.001) the agreement between methods without reducing the sensitivity to changes induced by tilting. Only the new method could provide estimates in all 2,584 files tested. Interpretation The new method intercepts the mean information of all other methods better than any other method, hence providing a simple, easy to standardize (no mathematical constraints) and yet robust and reliable BRS estimate. High-pass filtering markedly improves the agreement of all methods, without loss of sensitivity, and could be routinely used in clinical trials, to provide comparable BRS estimates.