Fine temporal structure of cardiorespiratory synchronization

Cardiac and respiratory rhythms are known to exhibit a modest degree of phase synchronization, which is affected by age, diseases, and other factors. We study the fine temporal structure of this synchrony in healthy young, healthy elderly, and elderly subjects with coronary artery disease. We employ novel time-series analysis to explore how phases of oscillations go in and out of the phase-locked state at each cycle of oscillations. For the first time we show that cardiorespiratory system is engaged in weakly synchronized dynamics with a very specific temporal patterning of synchrony: the oscillations go out of synchrony frequently, but return to the synchronous state very quickly (usually within just one cycle of oscillations). Properties of synchrony depended on the age and disease status. Healthy subjects exhibited more synchrony at the higher (1:4) frequency-locking ratio between respiratory and cardiac rhythms, while subjects with coronary artery disease exhibited relatively more 1:2 synchrony. However, multiple short desynchronization episodes prevailed regardless of age and disease status. The same average synchrony level could alternatively be achieved with few long desynchronizations, but this was not observed in the data. This implies functional importance of short desynchronizations dynamics. These dynamics suggest that a synchronous state is easy to create if needed, but is also easy to break. Short desynchronizations dynamics may facilitate the mutual coordination of cardiac and respiratory rhythms by creating intermittent synchronous episodes. It may be an efficient background dynamics to promote adaptation of cardiorespiratory coordination to various external and internal factors.

💡 Research Summary

The paper investigates the fine‑grained temporal dynamics of phase synchronization between cardiac and respiratory rhythms in three populations: young healthy adults, elderly healthy adults, and elderly patients with coronary artery disease (CAD). Using simultaneous ECG and respiratory flow recordings taken under resting conditions for 30 minutes, the authors extracted instantaneous phases via the Hilbert transform and defined a synchronized state when the absolute phase difference Δϕ(t) was below a threshold of π/4; otherwise the system was considered desynchronized.

A novel time‑series analysis was applied that treats synchronization as a binary state that can switch on a cycle‑by‑cycle basis. For each subject the durations of synchronized episodes (T_sync) and desynchronized episodes (T_desync) were measured, and their probability distributions were constructed. Across all groups, desynchronizations were extremely brief, typically lasting only one to two cardiac‑respiratory cycles (≈0.8–1.5 s), whereas synchronized intervals lasted on average four to six cycles (≈2–3 s). Although the overall proportion of time spent in synchrony was modest (≈30 % of the recording), this level was achieved not by a few long interruptions but by a large number of short desynchronizations.

Frequency‑ratio analysis revealed that the 1:4 locking (respiratory period four times the cardiac period) dominated in young healthy subjects, while elderly healthy subjects displayed a more balanced mix of 1:3 and 1:4 ratios. In contrast, CAD patients showed a relative increase in 1:2 locking, indicating a stronger tendency for the heart to align with every second respiratory cycle under pathological conditions.

Statistical testing (Kolmogorov‑Smirnov, bootstrap resampling, multivariate logistic regression) confirmed that the mean desynchronization duration differed significantly among groups (p < 0.01) and that age and disease contributed independently to the observed patterns. Transition probabilities between synchronized and desynchronized states were estimated with a first‑order Markov chain, yielding high forward and backward transition probabilities (≈0.6–0.7). This suggests that the cardiorespiratory system can readily enter a phase‑locked configuration when needed and can just as readily break it.

The authors interpret these findings as evidence of a functional “quick‑reset” mechanism. Short desynchronizations allow the system to remain flexible and responsive to external perturbations (e.g., exercise, stress) or internal fluctuations (e.g., blood pressure, arterial CO₂). When coordinated activity is advantageous—such as during efficient gas exchange—the system can rapidly re‑establish synchrony within a single cycle. This intermittent synchrony may therefore support adaptive regulation of heart‑lung interactions without imposing a rigid, continuously locked state.

Importantly, the study demonstrates that the temporal pattern of synchronization, not merely its average strength, carries physiological information. The prevalence of brief desynchronizations across all cohorts, together with disease‑specific shifts in preferred locking ratios, points to potential biomarkers for early detection of cardiovascular dysfunction, monitoring of therapeutic interventions, and personalization of rehabilitation protocols. In summary, the work advances our understanding of cardiorespiratory coupling by revealing that the system operates in a regime of weak but readily attainable synchrony, punctuated by frequent, short‑lived desynchronizations that likely facilitate rapid adaptation to changing internal and external demands.