Clinical Research Office. A partnership between Sheffield Teaching Hospitals NHS Foundation Trust and the University of Sheffield

CAOS Sleep Study



‘Cardiac Arrhythmogenesis in Sleep Apnoea’ (CAOS)

The aim of the study was to investigate the how common abnormal heart rhythms are in people with sleep apnoea. We also want to understand how heart rhythms might be linked to sleep apnoea.

Main results

104 participants were recruited into the study across two centers (Sheffield 79, Liverpool 25). There were 65 males and 39 females with an average age of 55 years.

86% of participants had sleep apnoea (OSA) with 27% mild, 23% moderate and 36% severe.


78% of participants had evidence of abnormal heart rhythm. This group had sleep apnoea that was worse and spent more of the night at lower oxygen levels. This measure is called the T90 score. Those with a worse T90 score seem to have a higher baseline heart rate and increased arrhythmia severity.

It is clear from this study that abnormal heart rhythms are much more common in patients with sleep apnoea than in the general population.

Preliminary results indicate that it is the average night-time oxygen level (T90 score) that dictates the risk of abnormal heart rhythm rather than the traditional severity scale which is called AHI (Apnoea-Hypopnoea Index) which we have always used to assess severity and make decisions around treatment. This is useful information that may guide how we discuss risk with patients in the future.

Future work

We are continuing to analyse the results in more detail. We have huge amounts of data from the three sleep studies completed and data from the activity watch and heart monitor. We plan to analyse the sleep tracking data simultaneously with the cardiac monitoring data. This is to get more granular detail about different parts of the night. This linked time-series data requires computing power that can only be made possible by the use of machine learning.

Once this is complete, we hope to identify a group of particularly vulnerable patients who are at increased risk. We would then like to run a further research study to follow at risk patients prospectively with or without the use of protective cardiac medications.

We presented our high-level results at the British Thoracic society meeting and we plan to publish a number of research papers with more detailed results, as well as developing future studies. We will be updating our website with more information as more data analyses is undertaken and published.

Dr Sriram Iyer

CAOS Chief Investigator

Consultant Physician

Sheffield Teaching Hospitals