Article Text

Download PDFPDF

Effect of CPAP on cardiovascular events in minimally symptomatic OSA: long-term follow-up of the MOSAIC randomised controlled trial
  1. Ivan Tang1,
  2. Chris D Turnbull2,3,
  3. Dushendree Sen2,
  4. Sonya Craig4,
  5. Malcolm Kohler5 and
  6. John R Stradling3
  1. 1Department of Anaesthesia and Intensive Care, Milton Keynes University Hospital, Milton Keynes, UK
  2. 2Oxford Centre for Respiratory Medicine, Oxford University Hospitals NHS Foundation Trust, Oxford, Oxfordshire, UK
  3. 3NIHR Biomedical Research Centre Oxford, University of Oxford, Oxford, Oxfordshire, UK
  4. 4Liverpool Sleep and Ventilation Centre, University Hospital Aintree, Liverpool, UK
  5. 5Centre for Interdisciplinary Sleep Research, University of Zurich, Zurich, Switzerland
  1. Correspondence to Dr Chris D Turnbull; christopher.turnbull{at}


The effect of continuous positive airway pressure (CPAP) on cardiovascular events is uncertain in minimally symptomatic obstructive sleep apnoea. Previous 2-year follow-up data from the Multicentre Obstructive Sleep Apnoea Intervention Cardiovascular (MOSAIC) trial showed a marginal reduction in cardiovascular events with CPAP therapy. We now present long-term MOSAIC study follow-up data. Median (first quartile, third quartile) follow-up was 5.0 (2.2, 5.0) and 3.7 (1.5, 5.0) years for CPAP and standard care, respectively. Compared to standard care, CPAP had no statistically significant effect on the risk of cardiovascular events (HR=0.83, p=0.54, 95% CI 0.46–1.51).

  • sleep apnoea
  • systemic disease and lungs

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The Multicentre Obstructive Sleep Apnoea Intervention Cardiovascular (MOSAIC) trial was a multicentre, randomised controlled trial investigating the effect of continuous positive airway pressure (CPAP) treatment on calculated cardiovascular risk and daytime sleepiness in patients with minimally symptomatic obstructive sleep apnoea (OSA). Although CPAP improved daytime sleepiness, it did not improve calculated cardiovascular risk at 6 months.1

Two-year follow-up data of the Oxford patients found a statistically significant but marginal reduction (absolute reduction 9.6%, 95% CI 0% to 19%, p=0.05) in new cardiovascular events in the CPAP group compared with standard care.2 This contrasts with the Sleep Apnoea Cardiovascular events (SAVE) Study which showed no improvement in new cardiovascular events following CPAP therapy.3

Despite evidence from observational studies, improvement in cardiovascular events resulting from CPAP in OSA remains unendorsed by randomised controlled trials.4


We report long-term follow-up from all patients from a single centre of the MOSAIC trial. The MOSAIC trial was a multicentre, randomised controlled parallel, hospital trial in patients with confirmed OSA (oxygen desaturation index >7.5/hour) but insufficient symptoms to mandate CPAP therapy.1

Patients were randomised to standard care or auto-adjusting CPAP therapy. The Oxford cohort was followed-up for a maximum of 5 years, and data were collected on new cardiovascular events. Data were analysed on an intention-to-treat basis. We prospectively recorded the occurrence of new cardiovascular events, including those leading to death. Cardiovascular events included angina, myocardial infarction, hypertension, peripheral vascular disease, atrial fibrillation, cardiac failure, stroke, transient ischaemic attack, coronary intervention and coronary artery bypass graft. Event-free survival was analysed using the log-rank test and Cox proportional HRs with multivariate analyses adjusted for age, baseline cardiovascular risk score and body mass index (BMI). A per protocol analysis was carried out of those patients who were randomised to CPAP who had usage of >4 hour/night at last follow-up.

At 2-year follow-up 9% and 4% in the standard care and CPAP arm groups experienced new events per year of follow-up.2 In order not to miss similar difference event rate over 5 years, with 80% power and two-sided alpha of 0.05, 120 patients were needed.


All 191 patients from a single centre are included in this follow-up analysis, with 94 initially randomised to the CPAP arm and 97 to the standard care arm. Patient characteristics are displayed in table 1. Forty-three patients (46%) randomised to CPAP crossed over and stopped CPAP. Those randomised to CPAP and still using CPAP had good mean adherence, with median of mean usage of 5:51 hours:min/night (Interquartile range 3:22–6:58) at the end of the follow-up period; and 10 standard care subjects (10%) had were using CPAP with median of mean usage of 3:32 hours:min/night (Interquartile range 1:00–6:55) by the end of the period.

Table 1

Patient characteristics by treatment group. Data are displayed as mean±standard deviation, median (first quartile, third quartile) or number (%)

There were 25 cardiovascular events in the CPAP group, compared with 32 in the standard care group. There was no significant difference in the univariate risk of cardiovascular events between CPAP and standard care (HR=0.83, p=0.54, 95% CI 0.46 to 1.51, figure 1). Multivariate adjustments for age, BMI and baseline cardiovascular risk score did not alter these results (HR=0.82, p=0.52, 95% CI 0.45 to 1.50). A per protocol analysis of the risk of cardiovascular events in patients with CPAP usage >4 hour/night was also not significant, compared with standard care (HR=1.3, p=0.46, 95% CI 0.61 to 2.94).

Figure 1

Kaplan-Meier curve showing the percentage of patients without the occurrence of a cardiovascular event by the number of follow-up days for standard care (dotted line) or CPAP solid line. Log rank p=0.54. Cox proportional HR=0.83 (p=0.54, 95% CI 0.46 to 1.51). CPAP, continuous positive airway pressure.


We have not shown CPAP to significantly impact cardiovascular event rates in minimally symptomatic OSA. This contrasts with our previous publication showing a reduction in cardiovascular events after 2 years of follow-up,2 hence the importance of reporting these longer term data. These results are in keeping with the much larger SAVE Study,3 which found that CPAP did not reduce cardiovascular events in patients screened for OSA following presentation to cardiovascular clinics.

Long-term follow-up in this study was limited by the small trial population, and thus wide confidence intervals of the estimates. However, given the marginal reduction in cardiovascular events observed at the 2 year follow-up, these data are important to report. The median length of follow-up was also longer in the CPAP arm as these patients were motivated to stay in touch to maintain their equipment. However, survival analyses take length of follow-up into account.

A theme consistent between MOSAIC, SAVE and the more recent Assessment of the Effect of CPAP on Energy and Vitality in Mild OSA (MERGE) trial, was the improvement of quality of life and symptoms with CPAP in even mild OSA.1–3 5 It is notable that the average usage of CPAP at the end of long-term follow-up, among those who did continue treatment was good (>5 hour/night). This demonstrates that there are patients who adhere to CPAP in the long-term despite minimal symptoms. However, even among these individuals with good long-term adherence to CPAP there was no suggestion of a reduction in cardiovascular events.


We would like to acknowledge support from the Oxford Health Services Research Committee and ResMed UK who supplied the CPAP machines for this trial. CT was funded and supported by an NIHR academic clinical lectureship. We would like to acknowledge the support of the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC). The views expressed are those of the authors and not necessarily of the NHS, the NIHR or the Department of Health.



  • IT and CDT are joint first authors.

  • Contributors SC, MK and JRS conceived the initial trial and were responsible for trial design. IT and DS were responsible for data collection. IT and CDT were responsible for data analysis, literature searches and preparing the first draft of the manuscript. All authors viewed, commented and took responsibility for the final manuscript.

  • Funding The British Heart Foundation—unrestricted project grant.

  • Competing interests JRS has done paid consulting work for Bayer and Resmed, outside of the scope of this work. CDT has done paid consulting work for Bayer, outside of the scope of this work.MK reports personal fees from Bayer, personal fees from Astra, personal fees from Boehringer, grants from Roche, grants and personal fees from GSK, personal fees from Novartis, personal fees from Philips Respironics, outside the submitted work.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; internally peer reviewed.

  • Data availability statement Data are available upon reasonable request including appropriate ethical approval. Qualified researchers may request access to patient level data and related study documents including the study protocol. Requests will be reviewed for scientific merit. Patient level data will be deidentified and study documents will be redacted to protect the privacy of trial participants. Please contact the corresponding author.