Non-invasive assessment of integrated cardiorespiratory network dynamics after physiological stress in humans

Visualizing the exchange of information between different components of the cardiorespiratory system is essential for understanding physiological control and is valuable for monitoring individuals in various clinical settings and extreme environments. Network physiology is a multidisciplinary field that explores the collective behaviour of physiological systems, offering a novel approach to understanding these complex interactions. The present study aimed to develop a non-invasive method to visualize the cardiorespiratory network under control conditions and after isolated or combined exposure to hypoxia, exercise and sleep-deprivation (total or partial). Using transfer entropy, a measure of causal relationships between time-series data, network maps were generated to illustrate information flow between 10 min parallel cardiorespiratory signals, namely, heart rate, respiratory rate, minute ventilation, tidal volume, capillary oxygen saturation ( $S_{p{O}_2}$ ), end-tidal O₂ and end-tidal CO₂ concentrations. Twenty-two healthy participants underwent assessments at rest and following normobaric hypoxia ( $F_{I{O}_2}$ : 0.12), moderate intensity cycling (100 W) and overnight sleep-deprivation both in isolation and in combination. The results showed that each stressor generated a distinct pattern of physiological information exchange. Increased connectivity within the networks was observed during exercise alone or when combined with other stressors. During exercise, heart rate emerged as the primary recipient of information, whereas SpO₂ served as the main disseminator. Hypoxia led to the engagement of $S_{p{O}_2}$ as a hub in the network. Sleep-deprivation was associated with a shift in the flow of information between the nodes during hypoxia. This non-invasive approach effectively maps cardiorespiratory interactions, offering potential for assessing integrated network dynamics in health and disease. KEY POINTS: Application of information theory principles can reveal significant bidirectional interactions between different components of the cardiorespiratory system using non-invasively recorded physiological signals. Increased connectivity within the cardiorespiratory networks was observed during moderate exercise alone or when combined with hypoxia and/or sleep-deprivation. During exercise, heart rate became the central node of the cardiorespiratory network for receiving information, whereas peripheral oxygen saturation acted as the main node that sent out information to other nodes. Acute hypoxic challenge or sleep-deprivation was associated with distinctive pattern of information flow in the cardiorespiratory network. This novel network approach can be used to map the adaptive responses to physiological stressors such as exercise, hypoxia and sleep-deprivation.