Sleep-disordered breathing and diabetes mellitus: a deadly duo

Diabetes mellitus type 2 (DM) is one of the most common chronic disorders in the developed world. In 2021, an estimated 529 million individuals worldwide were living with DM, with projections suggesting that this number could exceed 1.31 billion by 2050, underscoring DM as one of the most rapidly escalating global health issues.¹ Cardiovascular disease (CVD), including coronary artery disease, cerebrovascular events, heart failure, and peripheral arterial disease, are the leading complications associated with DM and represent the primary contributors to morbidity and mortality within this population.² Individuals with DM exhibit a 2- to 3-fold increased adjusted risk for CVD events compared to their counterparts without diabetes.³ Sleep-disordered breathing (SDB) is another prevalent chronic condition affecting up to a billion people worldwide.⁴ SDB has also been associated with an increased risk of developing cardiovascular and metabolic complications.⁵ Importantly, there is a reciprocal relationship between these 2 common conditions: there is an increased prevalence of SDB in patients with DM,⁶ and SDB may also serve as an independent risk factor for incident DM.⁷

In the present issue of the Journal of Clinical Sleep Medicine, Vichova et al⁸ tackle an important question, asking what is the additive risk of SDB on CVD and non-CVD mortality in patients with DM, who are already at high risk for these complications. Several studies have addressed this question in different cohorts and subpopulations. For example, a study from the SantOSA cohort in Chile demonstrated that patients with both moderate to severe obstructive sleep apnea and DM had a significantly higher risk of incident CVD mortality compared to those with either condition alone (hazard ratio 2.37, confidence interval 1.16–4.82).⁹ However, whether SDB in itself increases CVD mortality is not clear. Marin et al demonstrated that untreated severe SDB in men increased CVD mortality,¹⁰ but more recent data suggest no association between obstructive sleep apnea and increased CVD mortality.¹¹ Another study on moderate to severe SDB found an increased risk for all-cause and cancer mortality but not for CVD.¹²

Vichova et al used data from the Sleep Heart Health Study, a United States multicenter prospective observational study representing 5,780 patients with polysomnography and mortality data. The study included 453 patients with DM and compared their risk of CVD-related and nonrelated mortality. The main findings presented demonstrate that patients with DM who also have moderate to severe SDB face a significantly increased risk of all-cause mortality. When examining more nuanced subclassifications, a more complex picture emerged. Specifically, moderate-severe SDB, as measured by traditional metrics such as a higher apnea-hypopnea index or oxygen desaturation index, was associated with increased risk for overall mortality in patients with DM who are on treatment, but did not affect CVD mortality or any mortality parameters in the DM or non-DM group. Similarly, the findings on the impact of oxygen saturation levels on mortality risk were complex. Those with average oxygen saturation levels above 91.4% showed markedly lower all-cause and CVD mortality risks compared to their counterparts with lower oxygen saturation. Multivariate regression analysis suggests that a higher oxygen saturation is associated with a lower risk for CVD and all-cause mortality. The overall hypoxic burden, assessed as the cumulative time with oxygen saturation < 90% during sleep, was also associated with higher all-cause mortality in both the DM and non-DM groups. While higher oxygen saturation levels generally correlated with lower mortality risks, oxygen levels above 94.7% were associated with increased non-CVD mortality only in patients with untreated DM. Although the authors could not identify a reason for this finding, it highlights the complexity of interactions between DM and SDB, suggesting that optimal management may require a delicate balance rather than a simple “more is better” approach. These findings mirror those from a French cohort of patients newly diagnosed with SDB, which demonstrated that hypoxic burden and overall nocturnal hypoxemia assessed by sleep time with oxygen saturation < 90% were the only predictors of major adverse CVD events.¹³

One of the most intriguing aspects of this study is the divergent impact of sleep architecture on mortality between diabetic and nondiabetic individuals. While reduced sleep efficiency and rapid eye movement sleep were associated with increased mortality in the nondiabetic population, these factors appeared to have no significant impact on mortality in diabetic patients. This disparity raises intriguing questions about the underlying mechanisms of sleep’s influence on health outcomes in different metabolic states.

Despite its strengths, the study leaves several questions unanswered. The mechanisms underlying the differential effects of sleep architecture on mortality in diabetic vs nondiabetic individuals remain unclear. Additionally, the study’s observational nature precludes definitive conclusions about causality. The study’s methodology, which adjusted hazard ratios for confounding factors such as age, sex, race, BMI, hypertension, smoking status, and history of CVD, enhanced the robustness of its findings. Nonetheless, the groups differed significantly in their baseline characteristics, making any direct comparison and causal inference problematic. A more comprehensive analysis considering dyslipidemia treatment status and severity, alongside adjustments for other critical risk factors such as adiposity and glycemic control² would have provided a more nuanced understanding of their roles in the relationship between SDB and CVD mortality in patients with DM. Furthermore, an explicit examination of sex-specific effects would have been beneficial, particularly in light of evidence suggesting that DM is associated with a stronger relative risk for certain CVD events in females compared to males.^14,15 This point is further highlighted by the relative paucity of data on the sex differences in SDB mortality risk, opening opportunities for future research. An additional open question is the lack of specification of non-CVD causes of mortality, particularly the risk of cancer-related mortality in DM and SDB, which would be important to examine in the future. Finally, integrating the socioeconomic status of participants into the analysis would have been advantageous. Geographic disparities exist in the trends of diabetes complications and mortality; while high-income countries have experienced declines in CVD complications and all-cause mortality, low- and middle-income countries have witnessed increases. Defining participants’ socioeconomic status could have elucidated its role in the relationship between DM, SDB, and mortality.¹⁶

These limitations, however, present opportunities for future research. While the specific effects of continuous positive airway pressure or other oxygen therapy were not directly assessed, the association between SDB and mortality suggests that effective SDB management may improve outcomes in this high-risk population. Prospective studies examining the impact of SDB treatment on mortality in diabetic patients are thus urgently needed. Furthermore, the complex association between oxygen levels, diabetes status, and mortality suggests that we still do not fully understand which factors contribute more to the pathophysiology, whether it is the lack of oxygen, glucose intolerance, or the interaction between the two. Mechanistic investigations attempting to disentangle this complicated link between SDB, diabetes, and cardiovascular risk could uncover new therapeutic targets.

In summary, the study by Vichova et al⁸ advances our understanding of the impact of SDB and DM on mortality, emphasizing the importance of screening for SDB in diabetic patients as a potential modifiable risk factor.

DISCLOSURE STATEMENT

Each author listed on the manuscript has seen and approved the submission of this version of the manuscript and takes full responsibility for the manuscript. The authors report no conflicts of interest.