Can postural OSA be usefully identified from its severity alone?

When obstructive sleep apnoea (OSA) does not occur throughout sleep, there must be factors influencing its presence or absence. These are most likely to be sleep stage, posture and presleep alcohol, among others. We hypothesised that as OSA severity increases, the likelihood of postural OSA (POSA) would also decrease.

Laboratory sleep studies of 39 patients with OSA were manually reviewed to calculate supine and non-supine oxygen desaturation indices (ODI). The usual definition for POSA was used, a ratio of supine to non-supine ODI of ≥2.

The mean age was 53.2 (SD 12.4) years, the body mass index was 35.0 (SD 8.9) kg/m² and 74% were male. The median supine ODI was 54.3 (IQR 25.7-73.5) and non-supine ODI was 18.7 (IQR 8.6-38.4). The overall prevalence of POSA was 56%. The prevalence of POSA for ODIs of <40 was 68%, and 35% if ≥40.

An ODI ≥40, compared with <40, halved the likelihood of POSA from 68% to 35%. Although there is clearly a relationship between overall ODI and POSA, it is not strong enough to diagnose an individual with POSA. However the relationship provides a useful way to prescreen trial subjects to enrich for POSA.