Introduction Obstructive sleep apnoea (OSA) has an established independent association with insulin resistance and type 2 diabetes mellitus (T2DM). However, there are few data examining this relationship in severely obese populations, wherein any detrimental effect of OSA on metabolic health may conceivably be drowned out by the impact of morbid obesity. We assessed the relationship of OSA severity and nocturnal hypoxaemia with metabolic health in a cohort of severely obese patients attending sleep units across Europe.

Methods We performed a cross-sectional analysis of 1,434 participants in the European Sleep Apnea Cohort (ESADA) study with a body mass index (BMI) ≥35 kg/m², using multivariate regression analysis to assess T2DM prevalence according to OSA severity indices. Patients with diabetes were identified by history and medication prescription, and by screening for undiagnosed diabetes with glycosylated haemoglobin (HbA1c) measurement. The relationship of OSA severity with glycaemic control was assessed in diabetic subjects. Multivariate linear regression and multivariate analysis of co-variance were used to examine the relationship of HbA1c levels with OSA severity in both diabetic and non-diabetic patients.

Results In a cohort of predominantly male (63.5%) and severely obese (mean BMI 40.3 kg/m²) individuals, 32.2% had T2DM. Although the likelihood of T2DM was significantly greater in the highest AHI quartile than the lowest (unadjusted OR 1.78; 95% CI: 1.29–2.47), this relationship lost significance following adjustment for anthropometric, demographic, and clinical factors (adjusted OR 1.21; 95% CI: 0.83–1.76; p = 0.33). However, severity of nocturnal hypoxaemia remained a predictor of T2DM prevalence despite adjustment for confounding factors (adjusted OR for most severe mean nocturnal SpO₂ quartile 2.23; 95% CI: 1.50–2.20; p < 0.001), as well as predicting the likelihood of poor diabetic control (adjusted OR 1.87; 95% CI: 1.06–3.30; p = 0.03). In further analyses, HbA1c levels were independently predicted by OSA severity indices and nocturnal hypoxaemia in both non-diabetic and diabetic subjects, while adjusted mean HbA1c levels were significantly higher in patients with more severe sleep disordered breathing.

Conclusion Metabolic health in severely obese sleep patients appears to be significantly worse in those subjects with more severe sleep disordered breathing, with a particularly strong relationship with the degree of nocturnal hypoxaemia.