In our cohort study, OSA participants with diabetes had a higher incidence of all end events during the median 42-month follow-up. After adjusting for a range of potential confounders, our study showed a trend of increased risk for MACE, hospitalisation for unstable angina and a composite of all events in OSA patients with type 2 diabetes. Subgroup analysis demonstrated that adjusted hazard ratios for MACE by diabetes were higher in obese and overweight females ≥ 70 years and patients with mild OSA.
OSA is the most common type of sleep apnoea. It is caused by intermittent upper airway obstruction during sleep, resulting in repeated oxygenated haemoglobin desaturation and sleep fragmentation[21]. Multiple mechanisms link OSA to CVD complications, including insulin resistance, oxidative stress, sympathetic activation, endothelial dysfunction and increased inflammation[22]. A previous study showed a strong correlation between OSA and cardiac metabolic syndrome. However, this was a retrospective study of young and middle-aged adults[23]. The RICCADSA study also confirmed that OSA was an independent risk factor for poor cardiovascular prognosis in patients with acute coronary syndrome[24]. A meta-analysis revealed that the risk of fatal or non-fatal cardiovascular events in OSA patients was 3 times higher than controls [25]. Another study concluded that the mortality of ST-segment elevation MI in OSA patients was lower than non-OSA patients as OSA could initiate the mechanism of ‘ischaemic preconditioning’ to protect the myocardium[15]. There is a growing amount of evidence that the evolution of OSA severity is related to a deterioration in blood glucose control[26–27]. Therefore, our study further investigated the impact of concomitant type 2 diabetes on the long-term risk of MACE in patients with OSA. Notably, our study provided significant findings as a multicentre OSA population-based study adjusted for several potential confounders with confirmed statistical significance for MACE between patients with and without diabetes.
OSA could activate numerous endothelial cells and inflammatory cells and result in endothelial dysfunction, a predictor for MACE[28]. A cohort study proved that severe OSA was associated with cardiovascular events[29]. However, a cross-sectional study confirmed that moderate-severe OSA had no effect on microvascular endothelial function, especially in patients with type 2 diabetes[11]. Statistically speaking, although our data showed no correlation between the evolution of OSA severity and MACE risk in patients with type 2 diabetes, the risk trend for MACE increased in mild OSA patients with type 2 diabetes, which is partly consistent with previous study findings. First, age may be a significant interference factor in the results of this study. Second, severe OSA may involve more effective self-protective mechanisms, such as excessive respiratory effort and/or increased respiratory frequency, compensating for hypoxia in the body to reduce MACE risk.
OSA and all-cause mortality were significantly associated with each other in the general population. A study found that intermittent hypoxia could have protective effects on the cardiovascular system in elderly patients with OSA, reducing the risk of cardiovascular death and all-cause mortality [14]. Our findings showed that type 2 diabetes was nominally associated with the incidence of all-cause mortality and fell short of statistical significance, possibly because 87.6% of OSA patients with diabetes in our study were in stable condition with no target organ damage. Even so, the potential impact of the complications of diabetes on all-cause mortality and cardiovascular death in OSA patients cannot be ignored, especially in clinical diagnosis and treatment. Edwards et al. demonstrated that the severity of hypoxia caused by OSA in elderly patients is lower than in young patients[30]. Our data showed that the risk of MACE in elderly OSA ≥ 70 years with concomitant diabetes was significantly higher than in patients below 70 years, possibly due to the complex symptoms in elderly patients and impaired hypoxia tolerance. This study revealed that type 2 diabetes was associated with a higher risk of MACE in overweight and obese patients with OSA, which is not in line with previous studies. However, the ‘obesity paradox phenomenon’ indicated that obese patients with cardiovascular disease had a better cardiovascular prognosis than non-obese patients[31]. It is essential to regulate the body mass index, especially in elderly OSA patients with concomitant diabetes.
Evidence reveals that OSA and type 2 diabetes are independent risk factors for cardiovascular disease [1]. Previous studies showed that patients with OSA had a higher risk of cardiovascular disease [32–33]. However, a prospective survey of an Asian population showed no correlation between OSA and cardiovascular disease[34]. Adderley and Subramanian suggested that prevalent diabetes or incident diabetes during the follow-up period showed a higher CVD risk in OSA patients. However, in this study, most subjects were young and middle-aged patients in the UK[1]. Our study found that the elderly OSA patients with diabetes had a higher risk of MACE, especially females. Therefore, the relationship between diabetes and cardiovascular disease in OSA patients is worthy of further research.
Our study has several strengths and a few limitations. First, we assessed the risk of CVD and all-cause mortality in the diabetes group and the non-diabetes group of OSA patients without including healthy controls. Second, a median follow-up period of 42 months may be insufficient for all end events development in this cohort. Although this was a multicentre prospective cohort study, the study population consisted of Chinese patients; hence, selection bias could occur. However, these limitations do not affect the value of our study.