OSA Estimates for the UK 2019

It is clear that obstructive sleep apnoea (OSA) is of a similar order as asthma. However a precise estimation of OSA prevalence depends on the way it is defined, like many diseases. Because some people can have a degree of OSA with no daytime symptoms at all, the main confusion over estimates is whether they include subjects with relevant daytime symptoms such as excessive sleepiness or not. Another major difference between estimates is the rather arbitrary selection of a threshold for defining whether OSA is present; often divided into mild, moderate or severe (mild is 5 to 15 events per hour of sleep, moderate is 15 to 30, and severe over 30). Hence there are wildly different estimates of prevalence, which makes it hard to estimate the true health burden.

A recent major review of OSA prevalence from Benjafield and colleagues, in Lancet Respiratory Medicine, has tried to estimate the prevalence of OSA across the world based on sleep studies, regardless of symptoms. This is based on many prevalence studies which they have tried to normalise to a current standard way of measuring apnoeas and hypopnoeas (the 2012 criteria published by the American Association of Sleep Medicine, AASM) but this is inevitably imprecise. This comprehensive review estimates that in the UK, among subjects aged between 30 and 70 (30 million out of a total population of 66 million), 5% have moderate or severe OSA on a sleep study (1.5 million people), and a further 20% have mild OSA, thus 25% in total have mild, moderate or severe OSA (8 million people).

There will be additional subjects below 30 and above 70 in these OSA severity groups to take into account. This is only an estimate and, as obesity levels rise, so will the prevalence of OSA; for example the estimate for the USA for moderate and severe OSA is 15%, and for mild, moderate and severe it is 33%. This also makes the point that estimates will vary depending on the population being considered within the UK. For example, on average, heavy goods vehicle drivers are older and have higher BMIs so that the prevalence of OSA in them will be higher than these figures and nearer those of the USA.

As stated above, a significant factor influencing estimates is whether associated symptoms are included in the definition, and whether these might be severe enough to benefit from treatment. There is only a loose correlation between the severity of OSA on a sleep study and daytime symptoms. This is partly, but not wholly, because there is inter-individual variation in degree to which OSA events disrupt sleep, and because the measured OSA severity can vary considerably from night to night. Having said that, on the whole, the worse the OSA on a sleep study, the more the symptoms; but even apparently mild OSA on a sleep study can produce sufficient symptoms for a patient to benefit from CPAP.

Thus it is very clear that we do not have precise estimates for either the prevalence of OSA, OSA with symptoms (OSA syndrome, or OSAS), and OSAS that would benefit from a treatment such as continuous positive airway pressure (CPAP).

In the UK there are about 9 million people over 70, and 27 million under 30. If we assume that only 5 million of these 36 million have the same moderate and severe prevalence as the 30 to 70 age group, then the total in this moderate and severe group is 1.5 million plus 0.25 million, 1.75 million. If we further assume that the estimate for numbers potentially benefitting from a treatment (such as CPAP) works out as roughly the same as the prevalence (5%) of OSA in the moderate and severe group (ie subjects in this smaller group not benefitting from treatment roughly equals those in the larger mild group who would benefit from treatment) then we have an estimate for UK patients potentially benefitting from treatment of 1.75 million.

A very rough recent industry estimate of UK patients on CPAP already was about 0.8 million. Thus fewer than half the estimated potential patients have been treated so far. Sleep units in the UK continue to see yearly rises in the number of patients being referred and treated with CPAP so that, if anything, this 1.75 million feels like an under-estimate.

Thus in conclusion the quoted estimates for the prevalence of any level of OSA (mild, moderate or severe) in the UK could be argued to be 25%, but that only about 5% (1.75 million) would be likely to benefit from CPAP treatment. Presumably a larger number might benefit from other less intrusive treatments such as a dental device or weight loss.

John Stradling September 2019