All Obstructive Sleep Apnea Events Are Not Created Equal: The Relationship between Event-related Hypoxemia and Physiologic Response

Abstract

Rationale: Obstructive sleep apnea (OSA) severity is typically assessed by the apnea-hypopnea index (AHI), a frequency-based metric that allocates equal weight to all respiratory events. However, more severe events may have a greater physiologic impact. Objectives: The purpose of this study was to determine whether the degree of event-related hypoxemia would be associated with the postevent physiologic response. Methods: Patients with OSA (AHI, ⩾5/h) from the multicenter Canadian Sleep and Circadian Network cohort were studied. Using mixed-effect linear regression, we examined associations between event-related hypoxic burden (HB_ev) assessed by the area under the event-related oxygen saturation recording with heart rate changes (ΔHR_ev), vasoconstriction (vasoconstriction burden [VCB_ev] assessed with photoplethysmography), and electroencephalographic responses (power ratio before and after events). Results: Polysomnographic recordings from 658 patients (median [interquartile range] age, 55.00 [45.00, 64.00] yr; AHI, 27.15 [14.90, 64.05] events/h; 42% female) were included in the analyses. HB_ev was associated with an increase in all physiologic responses after controlling for age, sex, body mass index, sleep stage, total sleep time, and study centers; for example, 1 standard deviation increase in HB_ev was associated with 0.21 [95% confidence interval, 0.2, 0.22], 0.08 [0.08, 0.09], and 0.22 [0.21, 0.23] standard deviation increases in ΔHR_ev, VCB_ev, and β-power ratio, respectively. Conclusions: Increased event-related hypoxic burden was associated with greater responses across a broad range of physiologic signals. Future metrics that incorporate information about the variability of these physiologic responses may have promise in providing a more nuanced assessment of OSA severity.

Keywords: apnea–hypopnea index; electroencephalogram; heart rate response; hypoxic burden; obstructive sleep apnea.

Figure 1.

Illustration of changes in the physiologic responses to respiratory events. First box: Airflow signal. Second box: oxygen desaturation associated with the event. Third box: heart rate response. Fourth box: changes in the EEG power of the b-frequency band. Fifth box: Changes in the blood volume and vasoconstriction following the respiratory event. EEG = electroencephalography; HR = heart rate; PPG = photoplethysmography; Sp_O2 = arterial oxygen saturation as measured by pulse oximetry.

Figure 2.

Flow diagram presenting ascertainment of the study samples. AHI = apnea–hypopnea index; CSCN = Canadian Sleep and Circadian Network; EDF = European data format; EEG = electroencephalography; HB_ev = event-related hypoxic burden; HR = heart rate; PPG = photoplethysmography.

Figure 3.

Level of event-related physiologic responses in quartiles of HB_ev (shown by mean and standard error). Physiologic responses are greater in higher quartiles of HB_ev. HB_ev = event-related hypoxic burden; ΔHR_ev = event-related heart rate response; VCB_ev = event-related vasoconstriction burden; VCD_ev = event-related vasoconstriction depth.

Figure 4.

Correlation between HB_ev and event-related physiologic responses. HB_ev = event-related hypoxic burden; ΔHR_ev = event-related heart rate response; VCB_ev = event-related vasoconstriction burden; VCD_ev = event-related vasoconstriction depth.