Respiratory event type and duration modulate PPG waveforms in OSA.

Objective.Obstructive sleep apnea (OSA) increases circulatory system load and risk of cardiovascular diseases (CVDs). Despite being biomarkers of vascular health, photoplethysmography (PPG) signal waveforms have not been systematically investigated in patients with OSA. Thus, we aimed to study the immediate effects of respiratory event characteristics on PPG waveforms and to examine potential patterns that could be associated with cardiovascular health.Approach.We retrospectively analyzed 520 clinical polysomnographies from suspected OSA patients, focusing on three PPG signal segments: 10 s before, during, and 10 s after respiratory events (n= 14442 in each segment). We determined PPG waveform parameters (full width at half maximum (FWHM) as well as differences in absolute and relative time (Δt, Δtrel) and amplitude (ΔA, and ΔArel) between systolic peak and dicrotic notch) within each segment and tested the effect of respiratory event type (apnea vs hypopnea) and duration (10-20 s, 20-30 s, ⩾30 s).Main results.Before and during respiratory events, longer respiratory events significantly (p< 0.001) increased ΔAreland FWHM, and decreased ΔAand Δtrel. These findings were significantly (p< 0.001) more pronounced during apneas than hypopneas. After the events, all PPG waveform parameters fluctuated strongly. Moreover, the type and duration of the respiratory event, pre-event values of PPG parameters, and heart rate were significantly (p< 0.001) associated with changes in the parameter values during the event.Significance.By analyzing PPG characteristics, we could enhance the evaluation of physiological stress caused by respiratory events in patients with OSA and therefore provide tools for the risk assessment of CVDs.