CPAP Does Not Reduce High-Sensitivity C-Reactive Protein in Patients With Coronary Artery Disease and Obstructive Sleep Apnea

Obstructive sleep apnea (OSA) is associated with an increase in high-sensitivity C-reactive protein (hs-CRP). Studies suggested that the degree of severity of OSA in obese patients with no known coronary artery disease correlates with higher levels of hs-CRP and that continuous positive airway pressure (CPAP) could reduce this inflammation marker. In this study we tested the hypothesis that CPAP therapy could also reduce hs-CRP in cardiac patients with multiple comorbidities and known OSA. Sixty-two consecutive patients were included in this study. All patients were referred for a sleep test because of clinical suspicion of OSA. Clinical variables, body mass index (BMI), hs-CRP, and lipid profile were obtained at the time of their referral and at 126.2 +/- 33.7 days followup. Thirty-four patients (group A) underwent CPAP therapy and 28 patients did not (group B). The linear regression of hs-CRP level on the severity of the apnea-hypopnea index (AHI) was significant (p = 0.05), but this significance is lost when ln(hs-CRP) was used (p = 0.263). Through analysis of covariance, ln(hs-CRP) was predicted by BMI (p = 0.000) (R-Sq = 46.2%). In group A, and despite a significant drop in the AHI with CPAP [median difference = 29.7 (-41.8, 22.2)], there were no significant differences in patients' BMI, lipid profile, or hs-CRP [median difference = 0.15 (-0.83, 0.64)] (p = 0.53) on followup. When both groups A and B were compared, they had matched BMI, lipids, ejection fraction, blood pressure, age, creatinine, awake 02 saturations, alcohol consumption, coronary artery disease, and baseline and followup hs-CRP despite significant differences in baseline AHI (37.65 vs 14.30, respectively, p = 0.000)' We conclude that the degree of OSA or CPAP treatment does not independently predict levels of ln(hs-CRP) in g cardiac outpatients when other clinical variables, BMI, and lipids are adjusted for. BMI remains the strongesl, independent predictor of hs-CRP in this patient population.