Causal Analysis Is Needed to Evaluate Cardiorespiratory Interaction Alterations in Postural Orthostatic Tachycardia Syndrome Patients

Beatrice Cairo1, Beatrice De Maria2, Vlasta Bari3, Francesca Gelpi3, Maura Minonzio4, Franca Barbic4, Laura Dalla Vecchia2, Raffaello Furlan4, Alberto Porta1
1Department of Biomedical Sciences for Health, University of Milan, 2IRCCS Istituti Clinici Scientifici Maugeri, Milan, 3Department of Cardiothoracic, Vascular Anesthesia and Intensive Care, IRCCS Policlinico San Donato, 4Department of Internal Medicine, IRCCS Humanitas Clinical and Research Center


Abstract

Cardiorespiratory interactions can be characterized using signal processing analysis tools applied to spontaneous variability of heart period and respiration. Among these tools, both noncausal and causal approaches have been proposed. Previous studies have also demonstrated that cardiorespiratory interactions are under autonomic nervous system control, with a sympathetic activation causing a decrease of the coupling strength. The aim of this study is to describe cardiorespiratory interactions in a population known to feature an exaggerated sympathetic response to an orthostatic stressor such as patients affected by postural orthostatic tachycardia syndrome (POTS). Twelve female POTS patients (age: 36±10 yrs) and 14 female healthy controls (age: 37±8 yrs) underwent electrocardiogram and respiratory movement recordings during supine rest and head-up tilt. Cardiorespiratory interactions were assessed via a traditional model-based noncausal approach, squared coherence (K2), and a model-based causal approach, transfer entropy (TE), applied on the beat-to-beat heart period variability and respiratory series. TE was found to be significantly decreased in POTS patients compared to healthy (H) subjects during head-up tilt (POTS: 0.08±0.04; H: 0.18±0.12, p = 0.013), while no significant difference was detected during rest (POTS: 0.11±0.07; H: 0.16±0.11). K2 did not present any relevant difference between the two groups either at rest (POTS: 0.55±0.31; H: 0.57±0.29) or during head-up tilt (POTS: 0.45±0.27; H: 0.61±0.27). Results seem to suggest that the relationship from heart period variability to respiration along the reverse causal direction compared to that from respiration to heart period variability could be a confounding factor in noncausal tools such as K2. Therefore, we conclude that causal approaches are more suitable than noncausal methods in evaluating the modification of the magnitude of cardiorespiratory interactions in POTS patients during orthostatic challenge, with possible future applications in post-acute COVID-19 patients exhibiting symptoms of dysautonomia.