Quantifying the Autonomic Nervous System Influence on Heart Rate Turbulence using Partial Least Squares Path Modeling

Heart rate turbulence (HRT) is a physiological phenomenon used for cardiac risk stratification. Its alteration or absence indicates a significantly increased likelihood of mortality. However, the influence of the autonomic nervous system (ANS) on HRT needs to be further investigated. We propose a cause-effect relationship model to quantify the influence of the ANS. A set of 481 Holter-monitor recordings from different medical conditions were used, from THEW· acute myocardial infarction, coronary artery disease and end-stage renal disease. We proposed to model the relationship between HRT and ANS using Partial Least Squares Path Modeling (PLS-PM), a method for structural equation modeling that allows analyzing the relationships between the observed data and the latent variables. HRT parameters were estimated on individual ventricular premature complex (VPC) tachograms. ANS was assessed by heart rate variability indices computed from 3-min before VPC tachograms. The data set was split into low-risk and high-risk subgroups. SDN N, P
LP
, TS and TO were the most relevant variables. In low-risk, ANS activity was negatively related to HRT, while correlation between coupling interval and HRT on high-risk depends on the pathology. PLS-PM suggests that the influence of physiological variables on HRT is broken on high-risk. Results of the model are in agreement with the baroreflex hypothesis.