Background: Slow wave sleep (SWS) has been theorized as reflecting a homeostatic sleep process and is considered a state of recuperation. SWS is reduced in obstructive sleep apnea (OSA) patients, but SWS has not been specifically studied in respiratory failure patients. The aim of this study is to investigate SWS in predominantly hypercapnic respiratory failure patients. Methods: We analyzed sleep and arterial blood gas records of all 97 respiratory failure patients who underwent polysomnography and bilevel non-invasive ventilation (NIV) treatment in our laboratory from 2008 to July 2009. We also analyzed 32 initial diagnostic study data from these 97 patients. Results: The 97 patients had an average age of 58±15 (SD) years. Total sleep time was 320.3±82.8 (SD) min of which 32.9%±15.4 (%) was spent in SWS. This high percentage SWS correlated positively with awake arterial CO2 pressure (PCO2) in both the 97 treatment studies (r=0.35, p=0.001) and the 32 initial diagnostic studies (r=0.40, p=0.025). The relationship was particularly apparent in patients with obesity hypoventilation syndrome or overlap syndrome. Statistical modelling identified three significant predictor variables for SWS across both diagnostic and NIV nights: PCO2, arousal index and female gender. Conclusions: Patients with respiratory failure have a high percentage of EEG assessed SWS which is in part determined by disease specific variables such as hypercapnia as well as by traditional SWS determinants such as sleep fragmentation and gender.
- Respiratory failure
- Slow wave sleep