Abstract
Study Objectives:
Sleep-disordered breathing and diabetes mellitus (DM) are often concomitant; however, data on the impact of sleep-disordered breathing on mortality in the population with diabetes remain scarce.
Methods:
The population from the Sleep Heart Health Study, a multicenter prospective observational study representing 5,780 patients with polysomnography and mortality data, including 453 patients with DM, was analyzed to assess the impact of sleep-disordered breathing variables and the presence of DM on all-cause, cardiovascular disease, and noncardiovascular disease associated mortality. Survival analysis and proportional hazard regression models were used to calculate the adjusted hazard ratios (aHRs) for mortality.
Results:
Patients with DM and the average oxygen saturation > 91.4% had significantly lower all-cause (aHR 0.52, confidence interval [CI] 0.34–0.80) and cardiovascular disease mortality risk (aHR 0.44, CI 0.22–0.87) as compared with patients with oxygen saturation below this value. Apnea-hypopnea index > 31 (aHR 1.58, CI 1.10–2.28) and oxygen desaturation index > 13.3 (aHR 1.58, CI 1.10–2.25) were associated with increased all-cause mortality in participants with DM on treatment. Sleep efficiency and proportion of rapid eye movement sleep did not have any impact on mortality in patients with DM and thus differed significantly from individuals without DM, where increased all-cause mortality was observed in those with sleep efficiency < 81.4% (aHR 0.77, CI 0.68–0.87) or rapid eye movement sleep < 14.9% (aHR 0.78, CI 0.68–0.89).
Conclusions:
Patients with diabetes on treatment and moderate to severe sleep-disordered breathing experience increased all-cause mortality. Reduced average oxygen saturation predicted both all-cause and cardiovascular death in the population with diabetes.
Citation:
Vichova T, Petras M, Waldauf P, Westlake K, Vimmerova-Lattova Z, Polak J. Sleep-disordered breathing increases mortality in patients with diabetes. J Clin Sleep Med. 2025;21(1):89–99.
Keywords: sleep-disordered breathing, diabetes mellitus, mortality, cardiovascular disease, oxygen saturation
BRIEF SUMMARY
Current Knowledge/Study Rationale: Sleep-disordered breathing is a risk factor for cardiovascular and noncardiovascular mortality in the general population, nevertheless, it remains unclear whether sleep-disordered breathing conveys an additional risk of death in patients with diabetes who are already affected by increased mortality due to diabetes and the diabetes-associated cardiovascular risk profile. The objective of this study was to investigate the impact of sleep-disordered breathing on all-cause, cardiovascular, and noncardiovascular disease mortality.
Study Impact: We confirmed that sleep-disordered breathing increases mortality in patients with diabetes. Moreover, the influence of individual parameters may vary between individuals with diabetes and those without the condition that highlights the importance of preventive, diagnostic, and therapeutic interventions for patients with diabetes.
INTRODUCTION
Sleep-disordered breathing (SDB) and diabetes mellitus (DM) are conditions that affect millions of people worldwide. Their prevalence is increasing, presenting a significant public health challenge.1–3 Diabetes increases the risk of cardiovascular mortality, particularly due to coronary artery disease, by 2–4 times compared to the population without diabetes.4 Cardiovascular mortality rates (MRs) in patients with diabetes remain higher than those in their nondiabetic counterparts, even after treatment for cardiovascular risk factors such as hypertension or dyslipidemia. Persistent cardiovascular risk has been attributed to the effects of complex pathophysiological processes, including hyperglycemia, chronic inflammation, and insulin resistance, which contribute to the progression of atherosclerosis and vascular damage.5 Multiple studies have identified SDB as an independent risk factor for cardiovascular and all-cause mortality.6
Patients with type 1 or 2 DM exhibit a higher prevalence of SDB than those without diabetes.7 Repetitive hypoxia-reoxygenation cycles accompanied by increased sympathetic output may initiate or aggravate oxidative stress, induce inflammatory responses, endothelial dysfunction, insulin resistance, and the formation of advanced glycation products, further worsening vascular damage. Disruption of sleep architecture reflected in reduced rapid eye movement (REM) sleep and diminished slow-wave sleep has been suggested as a risk factor for all-cause, cardiovascular, and noncancer-related mortality.8 Finally, excessive daytime sleepiness resulting from repeated arousal from sleep in patients with SDB has been linked to a small but significant proportion of motor vehicle accidents and deaths.9
While epidemiological studies have identified SDB as a risk factor for cardiovascular and noncardiovascular mortality in the general population,10,11 relevant data for patients diagnosed with diabetes are scarce, despite elevated cardiovascular and overall mortality in these patients. The primary objective of this study was to investigate the impact of SDB severity on all-cause, cardiovascular disease (CVD), and non-CVD mortality, specifically in patients with diabetes, considering confounding factors such as age, sex, race, body mass index, hypertension, current or previous smoking, and CVD history.
METHODS
The reporting of this study was in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology statement.12
Study population
The Sleep Heart Health Study is a multicenter prospective observational cohort study designed to examine the consequences of SDB. The details of the study design have been reported previously.13–15 The cohort consists of 5,804 males and females aged 40 years or older.
The definition of prevalent CVD was based on self-reported data. They included history of acute myocardial infarction, history of nonfatal coronary heart disease, stroke, angina, congestive heart failure, coronary artery bypass graft—events were combined into a variable labeled “pre-existing CVD.” CVD deaths included death from myocardial infarction, fatal stroke, any fatal coronary heart disease, and any fatal CVD since baseline. Prevalent DM was defined based on self-reported diabetes and the use of oral hypoglycemic medications and/or insulin.16 Type II polysomnography studies were conducted in all patients to determine the number of apnea and hypopnea episodes per hour. Apnea-hypopnea index (AHI) was defined as limitation of flow > 90% for at least 10 seconds with no oxygen desaturation threshold (with or without arousal), hypopnea was defined as reduction of flow > 30% together with ≥ 3% oxygen desaturation or with arousal lasting for at least 10 seconds. Oxygen desaturation index (ODI) was defined as number of oxygen desaturations ≥ 3% for at least 10 seconds per hour of sleep.
Sleep efficiency (the proportion of time spent asleep relative to the time spent in bed) was monitored. Sleep recordings were staged in epochs of 30 seconds and categorized as stage 1 sleep, stage 2 sleep, stage 3 sleep + stage 4 sleep, slow-wave sleep, REM sleep, or awake. The proportions of REM and slow-wave sleep to total sleep were assessed. SDB treatment was defined as use of continuous positive airway pressure, bilevel positive airway pressure, mouthpiece, or oxygen therapy.
For more details on demographic and sleep-related parameters please see methods in the supplemental material.
Statistical analysis
Categorical variables, presented as absolute numbers and percentages, were compared using the χ2 or Fisher exact test. Continuous variables were reported as mean and standard deviation, median, and interquartile range. Survival analysis was performed using Cox regression for the assessment of all-cause mortality, and Fine-Gray regression for competing risk analysis to calculate CVD and non-CVD mortality. Both regressions were applied to continuous variables of sleep parameters to identify possible associations with mortality. Subsequently, continuous variables were stratified into 20-quantiles (ventiles) each representing 5%. To minimize influence of confounding factors, HR were adjusted by age, sex, race, body mass index, hypertension, treatment of SDB, smoker status, and history of CVD.
The hazard ratio (HR) for each 20-quantile of sleep parameter between patients with higher and lower values of these parameters was calculated to find a cutoff value. This value was adopted if the upper limit of the 95% confidence interval (CI) was lower than 0.9 or the lower limit of the 95% CI was higher than 1.10.
The null hypotheses for HR noninferiority and nonsuperiority were established as ln(HR) > −0.1 and ln(HR) < 0.1, respectively. If these hypotheses were rejected, then the alternative hypothesis was accepted, suggesting inferiority or superiority was achieved.
Furthermore, MRs, expressed per 1,000 person-years (p-y), were calculated for each sleep parameter if a cutoff value was found.
All tests were 2-tailed, and the level of significance was set at 0.05. Statistical analyses and regressions were performed using Prism 10 (GraphPad Software, Inc., San Diego, California) and STATA/SE version 18 software (StatCorp, Lakeway Drive, Texas), respectively.
Role of the funding sources
Study funders had no role in the study design, analysis, and interpretation of data, writing of the paper, or decision to submit the paper for publication.
RESULTS
Baseline characteristics
The cohort consisted of 5,804 males and females. The analysis included 5,780 patients, of them 453 (7.8%) patients with DM. In total, 24 patients were excluded, 4 patients due to missing information regarding DM, and 20 patients due to follow-up shorter than 15 days or a complete absence of follow-up (see flowchart in the supplemental material).
There were more men (55.8% vs 46.9%) and African Americans (15.9% vs 8.3%) among the participants with DM compared to those without DM (Table 1). Patients with DM were older (mean age 68.4 vs 62.7 years), showed a higher prevalence of CVD (24.7% vs 10.7%), and had more severe SDB as assessed by higher AHI (22.6 vs 17.6), lower sleep efficiency (83.2% vs 85.3%), less REM sleep (18.6% vs 19.9%), and longer time spent with oxygen saturation (SpO2) < 90% (6.3% vs 3.3%) than the non-DM population (Table 2). Patients with DM had higher all-cause (47.2%) and CVD mortality (18.5%) in comparison to patients without DM (20.4% and 5.2%, respectively). The mean follow-up time for patients with DM was 9.5 years vs 11.2 years for individuals without DM. Complete data on baseline characteristics are reported in Table S1 and Table S2 in the supplemental material.
Table 1.
Baseline characteristics of the study population, categorical parameters.
| DM | Non-DM | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total | Death | Total | Death | ||||||
| All | CVD | Non-CVD | All | CVD | Non-CVD | ||||
| Total | 453 | 214 (47.2%) | 84 (18.5%) | 114 (25.2%) | 5,327 | 1,089 (20.4%) | 277 (5.2%) | 695 (13.0%) | |
| Sex | Female | 200 (44.2%) | 89 (41.6%) | 36 (42.9%) | 48 (42.1%) | 2,825 (53.0%) | 513 (47.1%) | 127 (45.8%) | 338 (48.6%) |
| Race | White | 356 (78.6%) | 172 (80.4%) | 65 (77.4%) | 95 (83.3%) | 4,545 (85.3%) | 958 (88.0%) | 245 (88.4%) | 623 (89.6%) |
| African-American | 72 (15.9%) | 33 (15.4%) | 17 (20.2%) | 12 (10.5%) | 440 (8.3%) | 111 (10.2%) | 31 (11.2%) | 56 (8.1%) | |
| Other | 25 (5.5%) | 9 (4.2%) | 2 (2.4%) | 7 (6.1%) | 342 (6.4%) | 20 (1.8%) | 1 (0.4%) | 16 (2.3%) | |
| Previous CVD | No | 302 (66.7%) | 129 (60.3%) | 55 (65.5%) | 74 (64.9%) | 4,039 (75.8%) | 709 (65.1%) | 163 (58.8%) | 546 (78.6%) |
| Yes | 112 (24.7%) | 69 (32.2%) | 29 (34.5%) | 40 (35.1%) | 571 (10.7%) | 263 (24.2%) | 114 (41.2%) | 149 (21.4%) | |
| Unknown | 39 (8.6%) | 16 (7.5%) | 0 (0.0%) | 0 (0.0%) | 717 (13.5%) | 117 (10.7%) | 0 (0.0%) | 0 (0.0%) | |
| SDB Treatment | No | 363 (80.1%) | 165 (77.1%) | 65 (77.4%) | 89 (78.1%) | 4,445 (83.4%) | 830 (76.2%) | 222 (80.1%) | 536 (77.1%) |
| Yes | 16 (3.5%) | 6 (2.8%) | 2 (2.4%) | 4 (3.5%) | 142 (2.7%) | 35 (3.2%) | 8 (2.9%) | 24 (3.5%) | |
| Unknown | 74 (16.3%) | 43 (20.1%) | 17 (20.2%) | 21 (18.4%) | 740 (13.9%) | 224 (20.6%) | 47 (17.0%) | 135 (19.4%) | |
| Smoker | No | 199 (43.9%) | 82 (38.3%) | 39 (46.4%) | 40 (35.1%) | 2,496 (46.9%) | 451 (41.4%) | 120 (43.3%) | 279 (40.1%) |
| Yes | 32 (7.1%) | 14 (6.5%) | 1 (1.2%) | 10 (8.8%) | 524 (9.8%) | 113 (10.4%) | 26 (9.4%) | 77 (11.1%) | |
| Ex | 218 (48.1%) | 116 (54.2%) | 43 (51.2%) | 64 (56.1%) | 2,273 (42.7%) | 522 (47.9%) | 130 (46.9%) | 338 (48.6%) | |
| Unknown | 4 (0.9%) | 2 (0.9%) | 1 (1.2%) | 0 (0.0%) | 34 (0.6%) | 3 (0.3%) | 1 (0.4%) | 1 (0.1%) | |
| Hypertension | No | 151 (33.3%) | 60 (28.0%) | 22 (26.2%) | 34 (29.8%) | 3,156 (59.2%) | 437 (40.1%) | 84 (30.3%) | 318 (45.8%) |
| Yes | 302 (66.7%) | 154 (72.0%) | 62 (73.8%) | 80 (70.2%) | 2,171 (40.8%) | 652 (59.9%) | 193 (69.7%) | 377 (54.2%) | |
| Dyslipidemia | No | 346 (76.4%) | 166 (77.6%) | 69 (82.1%) | 86 (75.4%) | 4,722 (88.6%) | 955 (87.7%) | 229 (82.7%) | 619 (89.1%) |
| Yes | 106 (23.4%) | 47 (22.0%) | 14 (16.7%) | 28 (24.6%) | 592 (11.1%) | 132 (12.1%) | 47 (17.0%) | 76 (10.9%) | |
| Unknown | 1 (0.2%) | 1 (0.5%) | 1 (1.2%) | 0 (0.0%) | 13 (0.2%) | 2 (0.2%) | 1 (0.4%) | 0 (0.0%) | |
Categorical variables are expressed as counts and percentage (%). CVD = cardiovascular disease, DM = diabetes mellitus, Ex = former smoker, SDB = sleep-disordered breathing.
Table 2.
Baseline characteristics of the study population, continuous parameters.
| Total | DM | Non-DM | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Death | Death | ||||||||
| No | All | CVD | Non-CVD | No | All | CVD | Non-CVD | ||
| Age (years) | 63.2 (11.2) | 64.3 (9.6) | 72.9 (8.5) | 75.4 (5.8) | 72.5 (8.8) | 60.2 (10.2) | 72.5 (9.5) | 75.7 (8.0) | 72.5 (9.3) |
| BMI (kg/m2) | 28.2 (5.1) | 31.0 (5.6) | 29.6 (5.4) | 29.1 (4.9) | 29.8 (5.6) | 28.1 (5.0) | 27.3 (4.9) | 27.1 (4.6) | 27.5 (5.0) |
| Waist (cm) | 97.0 (13.7) | 105.0 (13.7) | 103.5 (13.3) | 102.3 (14.1) | 103.9 (12.4) | 96.1 (13.6) | 97.5 (13.2) | 97.3 (12.9) | 98.0 (13.2) |
| Neck (cm) | 37.8 (4.2) | 40.2 (4.4) | 39.8 (4.1) | 39.4 (4.3) | 39.7 (3.8) | 37.6 (4.2) | 38.0 (4.1) | 37.9 (3.8) | 37.8 (4.1) |
| Follow-up (years) | 11.0 (3.1) | 12.1 (1.8) | 6.7 (3.5) | 6.4 (3.5) | 6.9 (3.4) | 12.2 (1.8) | 7.1 (3.3) | 7.0 (3.2) | 6.9 (3.2) |
| SpO2 (%) | 94.6 (2.0) | 94.1 (2.6) | 94.0 (2.5) | 93.8 (2.6) | 93.9 (2.4) | 94.8 (1.9) | 93.9 (2.2) | 94.0 (2.2) | 93.7 (2.2) |
| Time of SpO2 < 90% (%) | 3.5 (10.4) | 4.7 (13.6) | 8.0 (19.0) | 8.5 (21.3) | 8.0 (17.9) | 2.6 (8.1) | 5.8 (13.8) | 5.6 (13.8) | 6.1 (14.4) |
| REM sleep (%) | 19.8 (6.3) | 19.5 (6.2) | 17.5 (6.6) | 17.8 (6.6) | 17.3 (6.8) | 20.3 (6.1) | 18.3 (6.7) | 18.7 (7.0) | 18.3 (6.5) |
| Slow-wave sleep | 18.2 (11.9) | 17.2 (11.9) | 16.6 (13.4) | 17.9 (14.3) | 15.6 (12.7) | 18.4 (11.4) | 17.9 (13.2) | 16.9 (12.5) | 17.9 (13.2) |
| Sleep efficiency (%) | 82.8 (10.6) | 81.5 (10.4) | 78.7 (12.5) | 77.8 (13.2) | 79.1 (11.9) | 84.0 (9.8) | 79.1 (11.9) | 78.5 (11.8) | 79.4 (11.8) |
| AHI | 18.0 (16.1) | 21.5 (17.4) | 23.9 (18.2) | 23.8 (18.2) | 23.8 (17.9) | 16.9 (15.4) | 20.2 (17.4) | 20.6 (15.6) | 20.1 (17.8) |
| ODI | 8.5 (9.3) | 10.4 (10.1) | 11.4 (11.0) | 11.3 (11.4) | 11.3 (10.7) | 7.9 (8.9) | 9.6 (9.7) | 9.6 (8.6) | 9.5 (9.8) |
Continuous variables are expressed as mean (standard deviation). AHI = apnea-hypopnea index, BMI = body mass index, CVD = cardiovascular disease, DM = diabetes mellitus, ODI = oxygen desaturation index, REM = rapid eye movement, SpO2 = average oxygen saturation.
Mortality: impact of SDB and DM
In total, 1,303 deaths occurred over the follow-up period, 133 of which were not determined to be either CVD or non-CVD.
Multivariate regression analysis was employed to assess which of the sleep-related variables, evaluated as continuous parameters, have significant relation to all-cause, CVD and non-CVD mortality. Results of the analysis are reported in Table S3 in the supplemental material. Based on the revealed association between the sleep parameter and specific mortality, adjusted HRs (aHRs) were established for continuous variables stratified into ventiles to determine the effect size cutoff in the appropriate study population. The MRs, including specific cutoff values, are reported in Table 3.
Table 3.
Hazard ratios, including cutoff values and mortality rates for all-cause and specific mortality, determined for populations of patients with DM (all and those on treatment) and patients without DM for individual sleep parameters.
| Sleep Parameter | Death | Population | Cutoff | MR (95% CI) for ≤ Cutoff* | MR (95% CI) for > Cutoff* | aHR (95% CI)** | P |
|---|---|---|---|---|---|---|---|
| SpO2 (%) | All | DM | 91.4 | 84.9 (59.0–122.2) | 45.5 (39.2–53.0) | 0.52 (0.34–0.80) | .003 |
| Treated DM | 91.4 | 91.1 (58.8–141.2) | 47.9 (40.1–57.2) | 0.44 (0.26–0.76) | .003 | ||
| Non-DM | 92.1 | 37.9 (33.0–43.6) | 15.7 (14.7–16.9) | 0.61 (0.52–0.71) | < .001 | ||
| CVD | DM | 91.4 | 35.7 (20.3–62.9) | 18.6 (14.7–23.5) | 0.44 (0.22–0.87) | .018 | |
| Treated DM | n.d. | ||||||
| Non-DM | n.d. | ||||||
| Non-CVD | DM | 94.7 | 22.4 (17.0–29.6) | 29.7 (22.7–38.7) | 1.87 (1.19–2.96) | .007 | |
| Treated DM | 90.2 | 64.0 (30.5–134.2) | 26.7 (21.2–33.7) | 0.44 (0.22–0.90) | .024 | ||
| Non-DM | 92.1 | 26.5 (22.4–31.3) | 9.8 (9.0–10.7) | 0.60 (0.49–0.75) | < .001 | ||
| Time of SpO2 < 90% (%) | All | DM | 9.1 | 46.3 (39.9–53.8) | 71.1 (52.3–96.5) | 1.86 (1.27–2.72) | .001 |
| Treated DM | 9.2 | 48.9 (41.0–58.3) | 78.9 (54.8–113.5) | 2.34 (1.50–3.67) | < .001 | ||
| Non-DM | 0.1 | 12.3 (11.0–13.7) | 22.6 (21.1–24.2) | 1.33 (1.16–1.52) | < .001 | ||
| REM sleep (%) | All | DM | n.d. | ||||
| Treated DM | n.d. | ||||||
| Non-DM | 14.9 | 26.9 (24.0–30.1) | 15.5 (14.5–16.7) | 0.78 (0.68–0.89) | < .001 | ||
| Sleep efficiency (%) | All | DM | n.d. | ||||
| Treated DM | n.d. | ||||||
| Non-DM | 81.4 | 28.2 (25.9–30.6) | 13.4 (12.3–14.6) | 0.77 (0.68–0.87) | < .001 | ||
| AHI | All | DM | n.d. | ||||
| Treated DM | 31.4 | 48.0 (39.7–58.0) | 68.3 (51.2–91.2) | 1.58 (1.10–2.28) | .014 | ||
| Non-DM | n.d. | ||||||
| ODI | All | DM | n.d. | ||||
| Treated DM | 13.3 | 48.9 (40.3–59.5) | 60.9 (46.3–80.1) | 1.58 (1.10–2.25) | .012 | ||
| Non-DM | n.d. |
*Expressed per 1,000 person-years. **aHR, including 95% CI with reference group having sleep parameter ≤ cutoff. AHI = apnea-hypopnea index, aHR = hazard ratio, CI = confidence interval, CVD = cardiovascular disease, DM = diabetes mellitus, MR = mortality rate, n.d. = not determined, REM = rapid eye movement, ODI = oxygen desaturation index, SpO2 = average oxygen saturation, SpO2 < 90% = percentage of sleep time spent with oxygen saturation below 90%.
It was found that patients with DM and > 91.4% SpO2 had a lower CVD MR of 18.6 (95% CI: 39.2–53.0) per 1,000 p-y compared to those with lower SpO2, who had an MR of 35.7 (95% CI: 20.3–62.9) per 1,000 p-y. The effect size, expressed by an aHR of 0.44 (95% CI: 0.22–0.87), confirmed a significantly higher risk of excess CVD deaths in patients with DM and < 91.4% SpO2. Although elevated all-cause MRs were found in these patients independently of average SpO2, patients with DM and > 91.4% exhibited a lower MR of 45.5 (95% CI: 39.2–53.0), as demonstrated by an aHR of 0.52 (95% CI: 0.34–0.80) (Table 3 and Figure 1). Moreover, the beneficial impact of > 91.4% SpO2 on all-cause mortality in these patients was further diminished by a higher non-CVD MR of 29.7 (95% CI: 22.7–38.7) per 1,000 p-y if they had SpO2 levels > 94.7%, as documented by an aHR of 1.87 (95% CI: 1.19–2.96). This effect was not observed in patients without DM or in patients with treated DM. All-cause or non-CVD mortality remained significantly lower in these participants with increased SpO2.
Figure 1. aHRs for all-cause or specific mortality in patients with SpO2 levels above vs below the cutoff values.
aHR = adjusted hazard ratio comparing the risk for the group having sleep parameter > cutoff value with reference group having sleep parameter ≤ cutoff, CI = confidence interval, CVD death = death from cardiovascular disease, DM = diabetes mellitus, non-CVD death = death from cause other than cardiovascular disease, SpO2 = oxygen saturation.
Similarly, the percentage of cumulative time with SpO2 below 90% during total sleep time significantly influenced all-cause mortality in participants independently of DM. The overall MRs in patients with DM were 71.1 (95% CI: 52.3–96.5) per 1,000 p-y and 46.3 (95% CI: 39.9–53.8) per 1,000 p-y for those above and below the cutoff value of 9.1%, respectively. The mortality risk was increased by 1.86, as confirmed by aHR.
Participants without DM who recorded no time spent with SpO2 below 90% achieved an overall MR of 12.3 (95% CI: 11.0–13.7) per 1,000 p-y. Conversely, those with at least 0.1% of time spent with lower SpO2 were 1.33 times more likely to experience mortality, with an MR of 22.6 (95% CI: 21.1–24.2) per 1,000 p-y.
Given that CVD mortality was not influenced by the proportion of time spent with SpO2 below 90%, the observed association with all-cause mortality was only related to non-CVD deaths.
The relationship between sleep architecture parameters, including the proportion of REM sleep and sleep efficiency, and mortality differed among patients with and without DM. While REM sleep and sleep efficiency did not exhibit any effect on mortality in patients with DM, their higher proportions in patients without DM decreased the all-cause MR 0.78-fold for > 14.9% REM sleep and 0.77-fold for 81.4% sleep efficiency (Table 3, Figure 2, and Figure S2 and Figure S3 in the supplemental material).
Figure 2. aHRs for all-cause mortality in patients with REM sleep and sleep efficiency above vs below the cutoff values.
AHI = apnea-hypopnea index, aHR = adjusted hazard ratio comparing the risk for the group having sleep parameter > cutoff value with reference group having sleep parameter ≤ cutoff, CI = confidence interval, CVD death = death from cardiovascular disease, DM = diabetes mellitus, non-CVD death = death from cause other than cardiovascular disease, ODI = oxygen desaturation index, REM = rapid eye movement.
Similar associations observed in all patients with DM were also confirmed in patients with DM undergoing treatment. Moreover, they demonstrated an increased risk of overall mortality if they had higher AHI or ODI. In patients with DM on treatment, the MR increased from 48.0 (95% CI: 39.7–58.0) per 1,000 p-y to 68.3 (95% CI: 51.2–91.2) per 1,000 p-y at AHI > 31.4, with the mortality risk being 1.58 times higher.
Furthermore, an ODI above 13.3 in these patients suggested an increased risk of all-cause deaths, with an aHR of 1.58 (95% CI: 1.10–2.25). Their MR of 60.9 (95% CI: 46.3–80.1) per 1,000 p-y was significantly higher than that of 48.9 (95% CI: 40.3–59.5) per 1,000 p-y observed in patients with a lower ODI (Table 3, Figure 3, and Figure S4 and Figure S5 in the supplemental material).
Figure 3. aHRs for all-cause mortality in patients with AHI and ODI levels above vs below the cutoff values.
AHI = apnea-hypopnea index, aHR = adjusted hazard ratio comparing the risk for the group having sleep parameter > cutoff value with reference group having sleep parameter ≤ cutoff. CI = confidence interval, CVD death = death from cardiovascular cause, DM = diabetes mellitus, non-CVD death = death from cause other than cardiovascular disease, ODI = oxygen desaturation index.
No association was revealed between all-cause or specific mortality and slow-wave sleep in all study populations.
Kaplan-Meier curves demonstrated a marked difference among patients with and without DM concerning specific sleep parameters and the identified cutoff values (Figure S6, Figure S7, Figure S8, and Figure S9 in the supplemental material).
DISCUSSION
This study investigated whether the presence of SDB affects mortality in patients diagnosed with diabetes. The impact of DM was observed through higher risk of excess all-cause and CVD death, which is in line with published data underpinning diabetes as a major risk factor of increased mortality.16,17 Diabetes increases the risk of hypoxic burden.7,10 Nevertheless, data on impact of concomitant diabetes and SDB on mortality are scant.10,18
Individual parameters associated with SDB available from the Sleep Heart Health Study database were analyzed in relation to presence of DM and mortality, which was categorized into all-cause, CVD and non-CVD mortality. The cutoff values for excess mortality risk for individual parameters were investigated, as the actual cutoffs in our study population may not precisely replicate the arbitrary reference values in the general population, due to its limited size or different characteristics of the studied subpopulations. For some of the parameters the cutoffs in relation to mortality may not be firmly set in the literature.8,19,20 The cutoff points for excess mortality that could be determined in this study such as the threshold for average SpO2 and AHI approached the arbitrarily used values for general population; for time spent with SpO2 < 90% they differed between participants with DM vs non-DM.
We demonstrated that moderate-to-severe SDB (expressed as AHI > 31 corresponding to severe SDB or ODI > 13.3, corresponding to moderate SDB) increased the all-cause but not CVD mortality in individuals with diabetes on treatment by 58%. The absence of observed impact on mortality in the remaining subgroups may be attributed to the fact that patients with DM treated with oral medication or insulin typically present with more advanced disease stages, thereby facing a higher excess mortality risk compared to those not requiring medication or non-DM participants. However, it can only be inferred that the patients with DM who are not receiving medication may not require pharmacological intervention, as the database does not provide objective indicators of diabetes control.
Although it has been proven that AHI is causally associated with the development of multiple CVD risk factors at the molecular level (eg, hypertension, dyslipidemia, and inflammation),21 and some smaller studies demonstrated that the coexistence of obstructive sleep apnea and DM increases the risk of CVD mortality,10,18 it seems that AHI is not a strong independent predictor of cardiovascular outcomes.22,23 ODI, similarly to AHI, describes intermittent hypoxia during sleep and has been linked to the CVD24,25 but has limited predictive value with regard to cardiovascular death.22
While both have been used as major indicators in SDB diagnosis and classification, neither of these metrics capture features of changes in SpO2, such as severity of episodic desaturations and the baseline SpO2 level from which they occur, for which cumulative effects may be clinically important.22 Additional variables may be needed for precise phenotyping to address mortality or comorbidity development.
For example, the average SpO2 represented a more sensitive predictor of CVD and all-cause mortality than AHI or ODI for patients with DM in this study, corresponding to the previous findings where mean and baseline SpO2 were associated with CVD outcomes22 whereas summative exposure to hypoxia (expressed as the hypoxic burden) closely predicted mortality.1–3 Patients with diabetes did not have lower average SpO2 than those without diabetes in the present study but spent more time with SpO2 < 90%,26 predicting all-cause mortality, which is consistent with previous observations.27
In contrast to a plausible relationship between increased mortality and lower SpO2, surprisingly, an increase in all-cause and non-CVD mortality was observed in those with DM and SpO2 above 94.7%. What seems to be a paradoxical finding, could be explained by other possible underlying causes contributing to excess noncardiovascular mortality in patients with diabetes and preserved SpO2. A recent study measured blood SpO2 in patients with diabetes and prediabetes. Persons with prediabetes and newly screened diabetes had lower SpO2s than persons with known diabetes and healthy controls.28 We can speculate that treatment of diabetes will improve SpO2, but other risks associated with diabetes persist and are reflected in increased mortality. Another explanation reflects the different pathophysiological phenotypes within obstructive sleep apnea. A low threshold for respiratory arousal has been described in a proportion of patients with sleep apnea.29 The higher arousability of these patients is then reflected in a shorter duration of respiratory events, which may not be associated with significant decreases in saturation.
The primary objective of the Sleep Heart Health Study was to assess cardiovascular events which explains why data elucidating noncardiovascular mortality are limited. However, the excess non-CVD mortality risk with higher SpO2 was detected only in individuals with DM in this study. In recent years decline in CVD mortality through improved prevention in those with diabetes and increase in other non-CVD causes (such as cancer-related mortality) has been observed, secondary to the survival advantage through CVD prevention.32 Thus, it remains probable but speculative whether excessive non-CVD mortality is linked to cancer,17 a hypothesis that needs to be validated in future studies designed to evaluate this primary endpoint. A higher incidence of car accidents (due to excessive daytime sleepiness) in patients with SDB has been previously reported.30,31
Sleep structure and duration32,33 should also not be neglected, as evidenced by studies showing that the proportion and/or duration of REM sleep is associated with all-cause mortality34 as well as cardiovascular and non-cancer-related mortality.8 In fact, the percentage of REM sleep was identified as the most important sleep stage variable predicting survival.8
The lack of association among sleep structure parameters such as proportion of REM sleep and sleep efficiency to mortality among patients with DM contrasted with non-DM participants. Diabetes is associated with different sleep architecture.35 The impact of sleep structure on mortality is multifaceted and the parameters such as REM sleep/sleep efficiency may not be sufficient to encompass the intricacy of the sleep structure disturbances in this subpopulation. A study including more complex sleep parameters may be required to ascertain the impact on mortality in those with DM.
In conclusion, our findings validate that SDB increases MRs among patients with diabetes. Furthermore, the influence of individual parameters may vary between individuals with diabetes and those without the condition. Considerations are warranted to establish optimal SDB screening or diagnostic protocols for the diabetic population, as well as to introduce adequate therapeutic and preventive measures once clinically significant SDB is present.
The inherent limitations of this study must be acknowledged. First, the presence of diabetes was based on self-reports and/or concomitant therapy for diabetes which may have underestimated the true prevalence of diabetes by 50%.26 Information on diabetes type, long-term compensation, duration, and presence of microvascular complications is unknown and was not included in the analysis. Although the present study was based on an analysis of the largest prospective population-based study cohort, the number of patients diagnosed with diabetes and experiencing CVD events or death is limited. Thus, negative results need to be interpreted with caution unless replicated in larger cohorts of patients with diabetes and SDB. Finally, no detailed information was available on the causes of non-CVD deaths (including cancer, infection, and suicide).
DISCLOSURE STATEMENT
All the authors have read and approved this manuscript. The paper was funded by the National Institute for Research of Metabolic and Cardiovascular Diseases (Programme EXCELES, ID Project No. LX22NPO5104)—Funded by the European Union—Next Generation EU and The Czech Ministry of Health Grant AZV NU21-01-00259. The Sleep Heart Health Study (SHHS) was supported by National Heart, Lung, and Blood Institute cooperative agreements U01HL53916 (University of California, Davis), U01HL53931 (New York University), U01HL53934 (University of Minnesota), U01HL53937 and U01HL64360 (Johns Hopkins University), U01HL53938 (University of Arizona), U01HL53940 (University of Washington), U01HL53941 (Boston University), and U01HL63463 (Case Western Reserve University). The National Sleep Research Resource was supported by the National Heart, Lung, and Blood Institute (R24 HL114473, 75N92019R002). The authors report no conflicts of interest.
Supplemental Materials
ABBREVIATIONS
- AHI
apnea-hypopnea index
- aHR
adjusted hazard ratio
- BMI
body mass index
- CI
confidence interval
- CVD
cardiovascular disease
- DM
diabetes mellitus
- HR
hazard ratio
- MR
mortality rate
- ODI
oxygen desaturation index
- p-y
person-years
- REM
rapid eye movement
- SDB
sleep-disordered breathing
- SpO2
oxygen saturation
REFERENCES
- 1. Saeedi P, Petersohn I, Salpea P, et al. ; IDF Diabetes Atlas Committee . Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: results from the international diabetes federation diabetes atlas, 9th edition . Diabetes Res Clin Pract. 2019. ; 157 : 107843 . [DOI] [PubMed] [Google Scholar]
- 2. Benjafield AV, Ayas NT, Eastwood PR, et al . Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis . Lancet Respir Med. 2019. ; 7 ( 8 ): 687 – 698 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. World Health Organization . World health statistics 2023: monitoring health for the SDGs, sustainable development goals . https://www.who.int/publications/i/item/9789240074323 . Updated May 19, 2023. Accessed May 28, 2023.
- 4. Einarson TR, Acs A, Ludwig C, Panton UH . Prevalence of cardiovascular disease in type 2 diabetes: a systematic literature review of scientific evidence from across the world in 2007–2017 . Cardiovasc Diabetol. 2018. ; 17 ( 1 ): 83 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dash S, Leiter LA . Residual cardiovascular risk among people with diabetes . Diabetes Obes Metab. 2019. ; 21 ( Suppl 1 ): 28 – 38 . [DOI] [PubMed] [Google Scholar]
- 6. Punjabi NM, Caffo BS, Goodwin JL, et al . Sleep-disordered breathing and mortality: a prospective cohort study . PLoS Med. 2009. ; 6 ( 8 ): e1000132 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Basille D, Timmerman M, Basille-Fantinato A, Al-Salameh A, Fendri S, Lalau JD . Screening for sleep-disordered breathing in people with type 1 diabetes by combining polysomnography with glucose variability assessment . Diabetes Res Clin Pract. 2022. ; 185 : 109786 . [DOI] [PubMed] [Google Scholar]
- 8. Leary EB, Watson KT, Ancoli-Israel S, et al . Association of rapid eye movement sleep with mortality in middle-aged and older adults . JAMA Neurol. 2020. ; 77 ( 10 ): 1241 – 1251 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sassani A, Findley LJ, Kryger M, Goldlust E, George C, Davidson TM . Reducing motor-vehicle collisions, costs, and fatalities by treating obstructive sleep apnea syndrome . Sleep. 2004. ; 27 ( 3 ): 453 – 458 . [DOI] [PubMed] [Google Scholar]
- 10. Labarca G, Dreyse J, Salas C, et al . Risk of mortality among patients with moderate to severe obstructive sleep apnea and diabetes mellitus: results from the SantOSA cohort . Sleep Breath. 2021. ; 25 ( 3 ): 1467 – 1475 . [DOI] [PubMed] [Google Scholar]
- 11. Marshall NS, Wong KKH, Cullen SRJ, Knuiman MW, Grunstein RR . Sleep apnea and 20-year follow-up for all-cause mortality, stroke, and cancer incidence and mortality in the Busselton health study cohort . J Clin Sleep Med. 2014. ; 10 ( 4 ): 355 – 362 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. von EE, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP ; STROBE Initiative . Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies . BMJ. 2007. ; 335 ( 7624 ): 806 – 808 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Zhang GQ, Cui L, Mueller R, et al . The national sleep research resource: towards a sleep data commons . J Am Med Inform Assoc. 2018. ; 25 ( 10 ): 1351 – 1358 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Quan SF, Howard BV, Iber C, et al . The sleep heart health study: design, rationale, and methods . Sleep. 1997. ; 20 ( 12 ): 1077 – 1085 . [PubMed] [Google Scholar]
- 15. Sleep Heart Health Study - 04 Dataset Introduction - Sleep Data - National Sleep Research Resource - NSRR . https://sleepdata.org/datasets/shhs/pages/04-dataset-introduction.md . Updated June 12, 2020. Accessed August 14, 2023.
- 16. Sarwar N, Gao P, Kondapally Seshasai SR, et al. ; Emerging Risk Factors Collaboration . Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies . Lancet. 2010. ; 375 ( 9733 ): 2215 – 2222 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Seshasai SRK, Kaptoge S, Thompson A, et al. ; Emerging Risk Factors Collaboration . Diabetes mellitus, fasting glucose, and risk of cause-specific death . N Engl J Med. 2011. ; 364 ( 9 ): 829 – 841 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Su X, Li JH, Gao Y, et al . Impact of obstructive sleep apnea complicated with type 2 diabetes on long-term cardiovascular risks and all-cause mortality in elderly patients . BMC Geriatr. 2021. ; 21 ( 1 ): 508 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Malhotra A, Ayappa I, Ayas N, et al . Metrics of sleep apnea severity: beyond the apnea-hypopnea index . Sleep. 2021. ; 44 ( 7 ): zsab030 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Boulos MI, Jairam T, Kendzerska T, Im J, Mekhael A, Murray BJ . Normal polysomnography parameters in healthy adults: a systematic review and meta-analysis . Lancet Respir Med. 2019. ; 7 ( 6 ): 533 – 543 . [DOI] [PubMed] [Google Scholar]
- 21. Briançon-Marjollet A, Weiszenstein M, Henri M, Thomas A, Godin-Ribuot D, Polak J . The impact of sleep disorders on glucose metabolism: endocrine and molecular mechanisms . Diabetol Metab Syndr. 2015. ; 7 ( 1 ): 25 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Linz D, Loffler KA, Sanders P, et al. ; SAVE (Sleep Apnea Cardiovascular Endpoints) Investigators . Low prognostic value of novel nocturnal metrics in patients with OSA and high cardiovascular event risk: post hoc analyses of the save study . Chest. 2020. ; 158 ( 6 ): 2621 – 2631 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Azarbarzin A, Sands SA, Stone KL, et al . The hypoxic burden of sleep apnoea predicts cardiovascular disease-related mortality: the osteoporotic fractures in men study and the sleep heart health study . Eur Heart J. 2019. ; 40 ( 14 ): 1149 – 1157 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Frangopoulos F, Nicolaou I, Zannetos S, Economou NT, Adamide T, Trakada G . Association between respiratory sleep indices and cardiovascular disease in sleep apnea—a community-based study in Cyprus . J Clin Med. 2020. ; 9 ( 8 ): 2475 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Wang L, Ou Q, Shan G, et al . Independent association between oxygen desaturation index and cardiovascular disease in non-sleepy sleep-disordered breathing subtype: a Chinese community-based study . Nat Sci Sleep. 2022. ; 14 : 1397 – 1406 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Resnick HE, Redline S, Shahar E, et al. ; Sleep Heart Health Study . Diabetes and sleep disturbances findings from the sleep heart health study . Diabetes Care. 2003. ; 26 ( 3 ): 702 – 709 . [DOI] [PubMed] [Google Scholar]
- 27. Baumert M, Immanuel SA, Stone KL, et al . Composition of nocturnal hypoxaemic burden and its prognostic value for cardiovascular mortality in older community-dwelling men . Eur Heart J. 2020. ; 41 ( 4 ): 533 – 541 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Laursen JC, Jepsen R, Bruun-Rasmussen NE, et al . Blood oxygen saturation is lower in persons with pre-diabetes and screen-detected diabetes compared with non-diabetic individuals: a population-based study of the Lolland-Falster health study cohort . Front Epidemiol. 2022. ; 2 : 1022342 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Eckert DJ, White DP, Jordan AS, Malhotra A, Wellman A . Defining phenotypic causes of obstructive sleep apnea: identification of novel therapeutic targets . Am J Respir Crit Care Med. 2013. ; 188 ( 8 ): 996 – 1004 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Tregear S, Reston J, Schoelles K, Phillips B . Obstructive sleep apnea and risk of motor vehicle crash: systematic review and meta-analysis . J Clin Sleep Med. 2009. ; 5 ( 6 ): 573 – 581 . [PMC free article] [PubMed] [Google Scholar]
- 31. Shiomi T, Arita AT, Sasanabe R, et al . Falling asleep while driving and automobile accidents among patients with obstructive sleep apnea-hypopnea syndrome . Psychiatry Clin Neurosci. 2002. ; 56 ( 3 ): 333 – 334 . [DOI] [PubMed] [Google Scholar]
- 32. de Andrés I, Garzón M, Reinoso-Suárez F . Functional anatomy of non-REM sleep . Front Neurol. 2011. ; 2 : 70 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Ma C, Pavlova M, Liu Y, et al . Probable REM sleep behavior disorder and risk of stroke: a prospective study . Neurology. 2017. ; 88 ( 19 ): 1849 – 1855 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Zhang J, Jin X, Li R, Gao Y, Li J, Wang G . Influence of rapid eye movement sleep on all-cause mortality: a community-based cohort study . Aging (Albany NY). 2019. ; 11 ( 5 ): 1580 – 1588 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Chen DM, Taporoski TP, Alexandria SJ, et al . Altered sleep architecture in diabetes and prediabetes: findings from the Baependi heart study . Sleep. 2024. ; 47 ( 1 ): zsad229 . [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.