Can the complete blood count be used as a reliable screening tool for obstructive sleep apnea?

Abstract

Since hypoxia increases erythropoietin production and inflammation, the complete blood count (CBC) has been proposed as an inexpensive alternative for obstructive sleep apnea (OSA) screening. The objective of this study was to determine whetheror not intermittent hypoxia and OSA severity, as measured by the mean oxygen saturation (SpO2) and apnea-hypopnea index (AHI), affect parameters measured by the CBC. This retrospective study included a total of 941 surgical patients who had a pre-operative home sleep study. The pre-operative CBC was extracted from the electronic patient records. Patients were stratified according to their AHI scores, into mild (AHI ≥ 5- < 15), moderate (AHI ≥ 15- < 30), and severe (AHI ≥ 30) OSA groups. There were 244 patients without OSA, 294 with mild, 223 with moderate, and 180 with severe OSA. Our analysis showed that hemoglobin (P = 0.010), hematocrit (P = 0.027), and basophils (P = 0.006) showed significant changes among the different severities of OSA. For mean SpO2, there were negative associations with body mass index (r = - 0.287; P < 0.001), age (r = - 0.077; P = 0.021), hemoglobin (r = - 0.208; P < 0.001), hematocrit (r = - 0.220; P < 0.001), red blood cells (r = - 0.107; P = 0.001), mean corpuscular volume (MCV) (r = - 0.159; P < 0.001), mean corpuscular hemoglobin (r = - 0.142; P < 0.001), and basophils (r = - 0.091; P = 0.007). All analyzed parameters remained within normal clinical range. Multivariable regression identified hemoglobin, MCV, and basophils to be independent predictors of mean SpO2 and AHI. Hemoglobin, MCV, and basophils were independently associated with intermittent hypoxia defined by mean SpO2 and AHI. Adding CBC parameters to other screening tools for OSA may have additional value due to its association with changes in mean SpO2.