Hematocrit levels in sleep apnea.

This study addresses the hypothesis that patients with obstructive sleep apnea, who exhibit recurrent episodes of oxygen desaturation at night, have higher hematocrit levels than nonapneic control subjects. We prospectively studied 624 patients referred to the sleep disorders center at St. Michael's Hospital because of suspicion of sleep apnea. All patients had nocturnal polysomnography and measurements of hematocrit level, hemoglobin value, WBC count, and platelet count. Smoking history and awake oxygen saturation (SaO2) was recorded in all of them. Nocturnal oxygenation was assessed using three indices: lowest nocturnal SaO2 (LoSaO2), mean nocturnal SaO2 (MnSaO2) and percent of total sleep time spent at SaO2 lower than 85 percent (TST85%). Patients with TST85% in the lowest quartile (TST85% = 0) had minimally lower hematocrit levels than patients with TST85% in the highest quartile (8 < or = TST85% < or = 90): 0.41 +/- 0.03 vs 0.40 +/- 0.02 in female subjects and 0.45 +/- 0.05 vs 0.43 +/- 0.05 in male subjects, respectively (p < 0.05). Multiple linear regression analysis revealed that MnSaO2, age, and pack-years of smoking were significant predictors of hematocrit level, but they accounted for only 9 percent of the variability in hematocrit level (multiple R2 = 0.087; p < 0.05). We conclude that intermittent nocturnal hypoxemia during episodes of apnea does not lead to clinical polycythemia, but is associated with minor elevations in hematocrit value. These small elevations are unlikely to be useful as markers of hypoxic stress associated with sleep apnea.