The influence of intermittent hypoxemia on platelet activation in obese patients with obstructive sleep apnea

Abstract

Objectives: Literature regarding platelet function in obstructive sleep apnea (OSA) has considerable limitations. Given the central role of platelets in atherothrombosis and the known cardiovascular risk of OSA, we hypothesized that OSA severity is predictive of platelet function, independent of known comorbidities.

Design: Obese subjects, without comorbidities, underwent overnight, in-lab polysomnography. The following morning, 5 biomarkers of platelet activation were measured by whole-blood flow cytometry at baseline and in response to agonists (no stimulation, stimulation with 5 μM ADP agonist, and stimulation with 20 μM ADP agonist): platelet surface P-selectin, activated glycoprotein (GP) IIb/IIIa, and GPIb receptor expression, platelet-monocyte aggregation (PMA) and platelet-neutrophil aggregation (PNA).

Results: Of the 77 subjects, 47 were diagnosed with OSA (median apnea-hypopnea index [AHI] of 24.7 ± 28.1/h in subjects with OSA and 3.0 ± 3.9/h in subjects without OSA, p < 0.001). The groups were matched for body mass index, with a mean body mass index of 40.3 ± 9.6 kg/m(2) in subjects with OSA and 38.9 ± 6.0 kg/m(2) in subjects without OSA (p = 0.48). A comparison of time spent with an oxygen saturation of less than 90% showed that subjects who had 1 minute or more of desaturation time per hour of sleep had lower GPIb fluorescence in circulating platelets, as compared with those subjects who had less than 1 minute of desaturation time per hour of sleep; similar findings were observed following 5 μM and 20 μM of ADP stimulation, as compared with control vehicle, suggesting higher levels of circulating platelet activity. In multivariate analyses, only nocturnal hypoxemia and female sex predicted agonist response. Platelet surface P-selectin, platelet surface-activated GPIIb/IIIa, PMA, and PNA were not significantly correlated with markers of OSA.

Conclusions: In obese patients with OSA, platelet activation is associated with greater levels of oxygen desaturation, compared with matched control subjects. Metrics other than AHI (e.g., hypoxemia) may determine OSA-related thrombotic risk.

Keywords: GPIb; Platelet activation; hypoxemia; lung; obesity; sleep apnea.

Figure 1

A greater degree of hypoxia is predictive of platelet activation Platelet surface glycoprotein Ib (GPIb) fluorescence intensity (expressed as arbitrary units) is decreased in subjects with greater desaturation (expressed as minutes of time spent with an SaO₂ < 90%, per hour of sleep). Data between the first and third quartiles are shown as a box, which is bisected by the median value; whiskers are used to represent the upper and lower limits. p values were determined using the Mann-Whitney test.

Figure 2

Obstructive sleep apnea (OSA) status does not predict platelet activation Subjects with an apnea-hypopnea index (AHI) < 10 events/h (n = 30) had glycoprotein (GP)Ib fluorescence intensity (expressed as arbitrary units) values similar to those of subjects with an AHI > 10 events/h. Data between the first and third quartiles are shown as a box, which is bisected by the median value; whiskers are used to represent the upper and lower limits. p values were determined using the Mann-Whitney test.

Figure 3

Shown below are the apnea-hypopnea index (AHI) and desaturation times of groups 1-3 Group 1: no obstructive sleep apnea (OSA) (n = 27), Group 2: OSA, but minimal desaturation (time spent at an SaO₂ of < 90% [T < 90%] was < 1 min/h of sleep, n = 29), and Group 3: OSA and desaturation (T < 90% was ≥ 1min/h of sleep, n = 13).

Figure 4

Subjects with obstructive sleep apnea (OSA) with desaturation had greater platelet reactivity Group 1 represents no OSA (n = 27), Group 2 represents OSA with minimal desaturation (time spent with an SaO₂ < 90%—represented as minutes of desaturation time < 90% divided by the hours of total sleep time [T < 90%]— was less than 1 min/h of sleep, n = 29), and Group 3 represents OSA and significant desaturation (T < 90% was ≥ 1 min/h of sleep, n = 13). Platelet-surface glycoprotein (GP)Ib fluorescence intensity (expressed as arbitrary units) was not decreased significantly in circulating platelets (no agonist) but was decreased upon stimulation with 0.5 μM ADP and 20 μM ADP, indicating increased platelet reactivity in Group 3. Data between the first and third quartiles are shown as a box, which is bisected by the median value; whiskers are used to represent the upper and lower limits. p values were determined using the Kruskall-Wallis test.