Introduction

In a subset of patients with obstructive sleep apnoea (OSA), all-cause mortality can be predicted by hypoxia-related parameters like the total time spent with nocturnal oxygen saturation <90% (T90) (Oldenburg et al. Eur Heart J 2016;37:1695-703). Both T90 and percentage of sleep time with oxygen saturation <90% (Tc90%) reflect the hypoxic burden related to sleep-time breathing.

Aims

To find the hypoxic metabolic threshold (HMT) in OSA, i.e. the Tc90% level, where the metabolome changes are at their peak.

Methods

In a prospective observational study of 51 patients with polysomnography-confirmed OSA, targeted metabolomic analysis of 187 metabolites in serum, obtained at 3 time points overnight, was performed with liquid chromatography-mass-spectrometry. Ranked general linear model for repeated measures was used to detect differences in metabolite contents between populations that remained below and equal to or higher than consecutively selected Tc90% cut-off values. HMT was determined by the maximum effect size assessed by the average Cohen?s f over all metabolites.

Results

The HMT was at Tc90% of 1.8 matching the Cohen?s f value of 0.165. At HMT, the highest Cohen?s f values were among sphingomyelins (0.322), followed by those in amino acids (0.160), lysophosphatidylcholines (0.157) and phosphatidylcholines (PCs) (0.155). Compared to patients with Tc90% being below HMT, in those with Tc90% being at HMT or higher, concentrations of 7 sphingomyelins, 2 PCs and one acylcarnitine were significantly increased.

Conclusions

In OSA, the largest hypoxia-related metabolomic changes occur at Tc90% of 1.8 with shifts in sphingomyelins? content contributing most to these changes.