Abstract
Background
Obstructive sleep apnea (OSA) has seen a rising prevalence and is closely linked with various cardiovascular diseases. Fluctuations in blood sugar levels, known as glycemic variability (GV), are linked to negative cardiovascular outcomes. The objective of this research is to explore how fluctuations in blood sugar levels affect the occurrence of atrial fibrillation (AF) and the rate of mortality during hospitalization in individuals with OSA.
Methods
This study conducted a retrospective analysis of patients diagnosed with OSA based on the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database (version 2.2) covering the 2008–2019 period. The relationship between GV and outcomes such as AF and in-hospital mortality was assessed through restricted cubic spline (RCS) models and logistic regression (LR), with AF and in-hospital mortality as the primary endpoints. The differences in the risk of AF and in-hospital mortality across various levels of GV were examined using Kaplan-Meier (K-M) survival analysis. Additionally, subgroup analyses were performed to further explore these correlations.
Results
The research involved 6,189 individuals, with a mean age of 64 years old, 36% of whom were women. Among the cohort, 176 patients (2.8%) died during hospitalization, and 673 patients developed AF in the hospital. The analysis using LR revealed a notable link between the GV index and both the risk of AF and in-hospital mortality among OSA patients. According to the RCS model, there was a clear dose-response relationship, revealing that higher values of the GV index corresponded to a heightened risk of AF and in-hospital mortality. Moreover, analysis using the K-M method showed that there were notable statistical variations in the risks of AF and mortality among OSA patients, stratified by quartiles of the GV index.
Conclusions
Patients with OSA who have an elevated GV index face a significantly higher risk of both AF and mortality during hospitalization, underscoring the importance of developing personalized glycemic management strategies to improve patient outcomes.
Keywords: Obstructive sleep apnea (OSA), atrial fibrillation (AF), glycemic variability (GV), Medical Information Mart for Intensive Care-IV database (MIMIC-IV database), mortality
Highlight box.
Key findings
• Higher glycemic variability (GV) is linked to increased risks of in-hospital atrial fibrillation (AF) and mortality in obstructive sleep apnea (OSA) patients. Analysis showed that higher GV levels correlate with greater risk, and patients in the highest GV quartile had significantly higher incidence of AF and deaths. These findings suggest GV as a key marker for adverse outcomes, emphasizing the importance of managing blood glucose fluctuations in these patients.
What is known and what is new?
• OSA is linked to higher risks of AF and cardiovascular complications.
• GV is known to contribute to adverse outcomes in critical illness and diabetes.
• This study shows, for the first time in OSA patients, that higher GV independently predicts in-hospital AF and mortality, highlighting its role as a novel risk marker and potential target for intervention.
What is the implication, and what should change now?
• GV should be monitored in OSA patients to assess cardiovascular risk.
• Managing glucose fluctuations may improve outcomes in this population.
• Future guidelines and interventions should consider GV as a target in OSA care.
Introduction
Obstructive sleep apnea (OSA) is a prevalent condition marked by recurrent breathing disruptions during slumber, resulting from an obstruction in the upper respiratory tract (1). Such disturbances can range from brief moments to extended periods, and may occur numerous times during the night, which can cause sleep disruption and lower the oxygen levels available to the body. Studies indicate that OSA affects approximately 20% to 30% of adults (2). OSA, as a common condition, has a considerable impact on an individual’s quality of life (QoL), e.g., excessive daytime sleepiness (EDS) may impair cognitive functioning, lead to difficulties in job performance, and increase the risk of accidents, especially in driving. Besides a decline in QoL, those affected by OSA could experience both short- and long-lasting effects, such as changes in cardiovascular health, metabolism, and cognitive function (3).
It is projected that by 2050, up to 12 million people in America will be affected by atrial fibrillation (AF), a widespread type of heart arrhythmia (4). AF not only leads to uncomfortable symptoms such as palpitations, dyspnea, and dizziness, but can also lead to serious complications such as stroke, which significantly affects the QoL and prognosis of patients (5).
Glycemic variability (GV) denotes the extent to which blood sugar levels vary over a period. GV is not only closely related to metabolic control in diabetic patients but is also strongly correlated with the likelihood of developing several complications, including cardiovascular diseases (6).
Recent research has shown a strong link between OSA and the onset of several cardiovascular diseases (7-9). Among them, the incidence of AF, a common arrhythmia, is significantly increased in OSA patients (10). GV, as an indicator of blood glucose fluctuation, has been recognized as possibly playing a significant role in OSA patients (11). Patients with OSA often experience insulin resistance and metabolic syndrome (12), which may lead to significant fluctuations in blood glucose levels. High GV not only affects metabolic control in diabetic patients, but may also increase the risk of AF by exacerbating electrophysiologic instability of the heart due to inflammatory responses and oxidative damage (13). However, the effect of GV on AF in OSA patients has not been reported. Therefore, we investigated the relationship between GV and the occurrence of AF as well as in-hospital mortality rates in OSA patients by utilizing the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database (version 2.2). This analysis not only offers insights into the underlying pathological mechanisms but also suggests novel ideas for clinical intervention, with the goal of improving cardiovascular health and QoL in OSA patients. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1678/rc).
Methods
Source of data
For this investigation, we analyzed data sourced from the MIMIC-IV database (version 2.2) (14), a substantial publicly available collection that includes more than 50,000 de-identified intensive care unit (ICU) patient records from Beth Israel Deaconess Medical Center (BIDMC) between the years 2008 and 2019. The BIDMC Institutional Review Board (IRB) approved the database under Protocol #2001P001699, allowing for a waiver of the informed consent requirement. This research did not necessitate obtaining written informed consent, as per the guidelines set by national laws and institutional policies. The author (L.L.) gained entry to the MIMIC-IV database (Record ID: 65303680). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.
Study population
The MIMIC-IV database was utilized to identify OSA patients through the application of International Classification of Diseases (ICD)-9 and ICD-10 classification codes. In our study, we focused solely on the initial hospitalization record for patients who had been admitted multiple times (n=13,997). The following patients were not included in the study: those who were under 18 years old (n=0), had less than 3 blood glucose measurements (n=7,643), less than 24 hours of hospital stay (n=4), and more than 30 days of hospital stay (n=161). The ultimate cohort comprised 6,189 subjects. Figure 1 illustrates the procedure for including patients in the study.
Figure 1.
Flow chart of patient selection. MIMIC-IV, Medical Information Mart for Intensive Care-IV.
Collection of data
Patient baseline features were obtained through the use of SQL alongside PostgreSQL (version 16). These characteristics included demographic data [age, sex, marital status, body mass index (BMI), and ethnicity], vital signs [heart rate (HR), respiratory rate (RR), systolic blood pressure (SBP), and diastolic blood pressure (DBP)], and severity of illness upon admission, as assessed by the acute physiology and chronic health evaluation III (APSIII), sequential organ failure assessment (SOFA), logistic organ dysfunction system (LODS), oxford acute severity of illness score (OASIS), and systemic inflammatory response syndrome (SIRS). Laboratory measurements included a wide range of parameters such as creatine kinase myocardial band (CKMB), international normalized ratio (INR), partial thromboplastin time (PTT), glucose, albumin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine, alkaline phosphatase (ALP), urea nitrogen (UN), lactate dehydrogenase (LD), total bilirubin (TB), and white blood cell (WBC). Other key markers included absolute monocyte count (AMC), red blood cell distribution width (RDW), red blood cell (RBC) count, hematocrit, absolute lymphocyte count (ALC), absolute basophil count (ABC), absolute eosinophil count (AEC), hemoglobin, platelets, absolute neutrophil count (ANC), potential of hydrogen (pH), central venous lactate concentration, total CO2, PCO2, PO2, and SPO2 peripheral. Additionally, information on warfarin use and patient comorbidities was recorded. For laboratory parameters measured multiple times, we used the first available result within 72 hours of admission. Measurements taken after the onset of AF were excluded. GV was calculated as the coefficient of variation (CV) for glucose, which is determined by dividing the standard deviation of blood glucose readings taken during the hospital stay by the average of all glucose measurements obtained repeatedly (15). Handling of missing data: missing data were managed using two approaches. We resorted to multiple imputations using the random forest method for variables that had missing data below 25%. In cases where over 25% of the data for a variable was absent, the missing entries were categorized separately and incorporated into the analysis as a binary variable.
Outcomes of the study
This research primarily aimed to assess the occurrence of AF during hospitalization and in-hospital mortality.
Statistical analysis
In cases of continuous variables that do not follow a normal distribution, we described the data using the median and quartiles, and employed the Mann-Whitney U test to compare groups. In the case of continuous variables that follow a normal distribution, we described them using the average and standard deviation, and employed the t-test to compare groups. To describe categorical variables, we employed frequency counts (as percentages) and utilized Fisher’s exact test or Pearson’s Chi-squared test to compare groups.
Odds ratios (OR) and their corresponding 95% confidence intervals (CIs) were estimated using logistic regression (LR) models, with various confounding factors taken into account [Model 1: unadjusted; Model 2: adjusted for age, sex, and race; Model 3: adjusted for a more comprehensive set of variables, including age, sex, race, hypertension (HP), diabetes mellitus (DM), heart failure (HF), BMI, coronary heart disease (CHD), CKMB, INR, albumin, ALT, AST, creatinine, ALP, UN, TB, WBC, AMC, RDW, RBC, hematocrit, hemoglobin, platelets, pH, PO2, HR, and the use of warfarin]. To address potential multicollinearity, the variance inflation factor (VIF) was analyzed for all variables, leading to the exclusion of any with a VIF exceeding 5 from the models.
The differences in the risk of AF and in-hospital mortality associated with different degrees of GV were analyzed using the Kaplan-Meier (K-M) survival technique. To investigate potential nonlinear associations of GV with the risks of AF and mortality during hospitalization, restricted cubic spline (RCS) diagrams were used. Furthermore, analyses of subgroups were performed to investigate the relationships between exposure and results in groups with varying traits, and the results were visualized using forest plots.
R software, version 4.4.1, was employed for all statistical analysis procedures. A P value from a two-tailed test that falls below 0.05 is deemed to reflect statistical significance.
Results
Baseline features
Our analysis involved 6,189 patients who met the criteria for inclusion. Within the study cohort, 176 patients died during hospitalization, while AF was observed in 673 patients during hospitalization.
The characteristics at baseline for patients with AF and those without are compared in Table 1. The AF cohort consisted of a greater number of male individuals, characterized by older age and a higher occurrence of CHD and HF, while they experienced fewer cases of HP and utilized warfarin less frequently. Laboratory tests revealed that individuals with AF showed increased levels of INR, PTT, Glu, Creatinine, UN, WBC, ANC, and GV, but decreased levels of DBP, ALT, AST, and RDW (all P<0.05).
Table 1. Patient characteristics (grouped by AF).
| Characteristic | Overall (n=6,189) | Without AF (n=5,516) | With AF (n=673) | P value |
|---|---|---|---|---|
| In-hospital deaths (rate) | 176 (2.8) | 101 (1.8) | 75 (11.0) | <0.001*** |
| Length of hospital stay (days) | 6.5 (4.3, 9.8) | 6.1 (4.2, 9.1) | 9.8 (6.6, 15.0) | <0.001*** |
| Follow-up time for AF (days) | 5.8 (3.8, 8.8) | 6.1 (4.2, 9.1) | 1.3 (0.1, 3.2) | <0.001*** |
| Gender | <0.001*** | |||
| Female | 2,216 (36.0) | 2,016 (37.0) | 200 (30.0) | |
| Male | 3,973 (64.0) | 3,500 (63.0) | 473 (70.0) | |
| Anchor_age (years) | 64 (55, 72) | 63 (54, 71) | 70 (62, 77) | <0.001*** |
| Marital_status | <0.001*** | |||
| Divorced | 526 (8.5) | 470 (8.5) | 56 (8.3) | |
| Married | 3,267 (53.0) | 2,895 (52.0) | 372 (55.0) | |
| Null | 224 (3.6) | 183 (3.3) | 41 (6.1) | |
| Single | 1,616 (26.0) | 1,485 (27.0) | 131 (19.0) | |
| Widowed | 556 (9.0) | 483 (8.8) | 73 (11.0) | |
| Race | <0.001*** | |||
| Asian | 90 (1.5) | 80 (1.5) | 10 (1.5) | |
| Black | 737 (12.0) | 695 (13.0) | 42 (6.2) | |
| Hispanic | 237 (3.8) | 223 (4.0) | 14 (2.1) | |
| Other | 673 (11.0) | 560 (10.0) | 113 (17.0) | |
| White | 4,452 (72.0) | 3,958 (72.0) | 494 (73.0) | |
| BMI (kg/m2) | <0.001*** | |||
| <18.5 | 39 (0.6) | 35 (0.6) | 4 (0.6) | |
| ≥30 | 3,110 (50.0) | 2,827 (51.0) | 283 (42) | |
| 18.5–24.9 | 463 (7.5) | 425 (7.7) | 38 (5.6) | |
| 25.0–29.9 | 1,028 (17.0) | 938 (17.0) | 90 (13.0) | |
| Missing | 1,549 (25.0) | 1,291 (23.0) | 258 (38.0) | |
| SBP (mmHg) | 130 (122, 140) | 130 (122, 140) | 131 (122, 140) | 0.20 |
| DBP (mmHg) | 75 (69, 80) | 75 (69, 80) | 74 (68, 78) | <0.001*** |
| HP | 2,937 (47.0) | 2,643 (48.0) | 294 (44.0) | 0.04* |
| CHD | 2,144 (35.0) | 1,825 (33.0) | 319 (47.0) | <0.001*** |
| DM | 2,740 (44.0) | 2,443 (44.0) | 297 (44.0) | >0.90 |
| HF | 2,042 (33.0) | 1,678 (30.0) | 364 (54.0) | <0.001*** |
| CKMB (ng/mL) | 0.70 | |||
| >10 | 290 (4.7) | 254 (4.6) | 36 (5.3) | |
| 0–10 | 1,526 (25.0) | 1,360 (25.0) | 166 (25.0) | |
| Missing | 4,373 (71.0) | 3,902 (71.0) | 471 (70.0) | |
| INR | 1.23 (1.10, 1.50) | 1.20 (1.10, 1.50) | 1.37 (1.20, 1.52) | <0.001*** |
| PTT (sec) | 32 (28, 37) | 32 (28, 37) | 33 (29, 36) | <0.001*** |
| Glu (mg/dL) | 127 (104, 162) | 125 (103, 163) | 135 (115, 154) | <0.001*** |
| Albumin (4.2 g/dL) | 0.50 | |||
| <3.5 | 1,165 (19.0) | 1,051 (19.0) | 114 (17.0) | |
| >5.2 | 2 (<0.1) | 2 (<0.1) | 0 (0.0) | |
| 3.5–5.2 | 1,246 (20.0) | 1,104 (20.0) | 142 (21.0) | |
| Missing | 3,776 (61.0) | 3,359 (61.0) | 417 (62.0) | |
| ALT (U/L) | 0.01* | |||
| >40 | 849 (14.0) | 776 (14.0) | 73 (11.0) | |
| 0–40 | 2,299 (37.0) | 2,061 (37.0) | 238 (35.0) | |
| Missing | 3,041 (49.0) | 2,679 (49.0) | 362 (54.0) | |
| AST (U/L) | 0.01* | |||
| >40 | 989 (16.0) | 880 (16.0) | 109 (16.0) | |
| 0–40 | 2,173 (35.0) | 1,970 (36.0) | 203 (30.0) | |
| Missing | 3,027 (49.0) | 2,666 (48.0) | 361 (54.0) | |
| Creatinine (mg/dL) | 1.00 (0.80, 1.40) | 1.00 (0.80, 1.40) | 1.20 (0.90, 1.49) | <0.001*** |
| ALP (U/L) | 0.20 | |||
| <40 | 113 (1.8) | 102 (1.8) | 11 (1.6) | |
| >130 | 542 (8.8) | 490 (8.9) | 52 (7.7) | |
| 40–130 | 2,460 (40.0) | 2,212 (40.0) | 248 (37.0) | |
| Missing | 3,074 (50.0) | 2,712 (49.0) | 362 (54.0) | |
| UN (mg/dL) | 19 (14, 28) | 18 (13, 27) | 21 (17, 31) | <0.001*** |
| LD (U/L) | 0.80 | |||
| <94 | 4 (<0.1) | 4 (<0.1) | 0 (0.0) | |
| >250 | 906 (15.0) | 804 (15.0) | 102 (15.0) | |
| 94–250 | 1,284 (21.0) | 1,152 (21.0) | 132 (20.0) | |
| Missing | 3,995 (65.0) | 3,556 (64.0) | 439 (65.0) | |
| TB (mg/dL) | 0.093 | |||
| >1.5 | 407 (6.6) | 366 (6.6) | 41 (6.1) | |
| 0–1.5 | 2,717 (44.0) | 2,445 (44.0) | 272 (40.0) | |
| Missing | 3,065 (50.0) | 2,705 (49.0) | 360 (53.0) | |
| WBC (K/μL) | 9.4 (7.0, 12.4) | 9.1 (6.9, 12.3) | 10.1 (8.4, 12.7) | <0.001*** |
| AMC (K/μL) | 0.20 | |||
| <0.2 | 111 (1.8) | 95 (1.7) | 16 (2.4) | |
| >0.8 | 506 (8.2) | 447 (8.1) | 59 (8.8) | |
| 0.2–0.8 | 735 (12.0) | 644 (12.0) | 91 (14.0) | |
| Missing | 4,837 (78.0) | 4,330 (78.0) | 507 (75.0) | |
| RDW (%) | 14.20 (13.40, 15.50) | 14.20 (13.30, 15.40) | 14.11 (13.60, 15.50) | 0.01* |
| RBC (m/μL) | 3.95 (3.43, 4.39) | 3.94 (3.42, 4.41) | 4.02 (3.59, 4.19) | 0.80 |
| Hematocrit (%) | 35.6 (31.1, 39.3) | 35.4 (31.0, 39.4) | 36.4 (32.3, 38.3) | 0.12 |
| ALC (K/μL) | 0.08 | |||
| <1.2 | 524 (8.5) | 469 (8.5) | 55 (8.2) | |
| >3.7 | 55 (0.9) | 49 (0.9) | 6 (0.9) | |
| 1.2–3.7 | 773 (12.0) | 668 (12.0) | 105 (16.0) | |
| Missing | 4,837 (78.0) | 4,330 (78.0) | 507 (75.0) | |
| ABC (K/μL) | 0.005** | |||
| <0.01 | 175 (2.8) | 147 (2.7) | 28 (4.2) | |
| >0.08 | 73 (1.2) | 58 (1.1) | 15 (2.2) | |
| 0.01–0.08 | 1,104 (18.0) | 981 (18.0) | 123 (18.0) | |
| Missing | 4,837 (78.0) | 4,330 (78.0) | 507 (75.0) | |
| AEC (K/μL) | 0.30 | |||
| <0.04 | 363 (5.9) | 319 (5.8) | 44 (6.5) | |
| >0.54 | 51 (0.8) | 46 (0.8) | 5 (0.7) | |
| 0.04–0.54 | 938 (15.0) | 821 (15.0) | 117 (17.0) | |
| Missing | 4,837 (78.0) | 4,330 (78.0) | 507 (75.0) | |
| Hemoglobin (g/dL) | 11.60 (10.10, 13.00) | 11.60 (10.10, 13.00) | 11.90 (10.50, 12.50) | 0.50 |
| Platelets (K/μL) | 211 (162, 260) | 210 (162, 263) | 212 (167, 235) | 0.059 |
| ANC (K/μL) | 32 (28, 37) | 32 (28, 37) | 33 (29, 36) | <0.001*** |
| pH | <0.001*** | |||
| <7.35 | 907 (15.0) | 757 (14.0) | 150 (22.0) | |
| >7.45 | 326 (5.3) | 283 (5.1) | 43 (6.4) | |
| 7.35–7.45 | 1,705 (28.0) | 1,470 (27.0) | 235 (35.0) | |
| Missing | 3,251 (53.0) | 3,006 (54.0) | 245 (36.0) | |
| Total CO2 (mEq/L) | <0.001*** | |||
| <21 | 219 (3.5) | 176 (3.2) | 43 (6.4) | |
| >30 | 623 (10.0) | 529 (9.6) | 94 (14.0) | |
| 21–30 | 2,006 (32.0) | 1,718 (31.0) | 288 (43.0) | |
| Missing | 3,341 (54.0) | 3,093 (56.0) | 248 (37.0) | |
| Lactate (mmol/L) | <0.001*** | |||
| <0.5 | 6 (<0.1) | 5 (<0.1) | 1 (0.1) | |
| >2 | 681 (11) | 574 (10.0) | 107 (16.0) | |
| 0.5–2 | 2,137 (35.0) | 1,832 (33.0) | 305 (45.0) | |
| Missing | 3,365 (54.0) | 3,105 (56.0) | 260 (39.0) | |
| PCO2 (mmHg) | <0.001*** | |||
| <35 | 339 (5.5) | 289 (5.2) | 50 (7.4) | |
| >45 | 1,203 (19.0) | 1,016 (18.0) | 187 (28.0) | |
| 35–45 | 1,304 (21.0) | 1,116 (20.0) | 188 (28.0) | |
| Missing | 3,343 (54.0) | 3,095 (56.0) | 248 (37.0) | |
| PO2 (mmHg) | <0.001*** | |||
| <85 | 1,004 (16.0) | 851 (15.0) | 153 (23.0) | |
| >105 | 1,544 (25.0) | 1,304 (24.0) | 240 (36.0) | |
| 85–105 | 298 (4.8) | 266 (4.8) | 32 (4.8) | |
| Missing | 3,343 (54.0) | 3,095 (56.0) | 248 (37.0) | |
| RR (bpm) | <0.001*** | |||
| <12 | 236 (3.8) | 202 (3.7) | 34 (5.1) | |
| >20 | 785 (13.0) | 616 (11.0) | 169 (25.0) | |
| 12–20 | 1,681 (27.0) | 1,357 (25.0) | 324 (48.0) | |
| Missing | 3,487 (56.0) | 3,341 (61.0) | 146 (22.0) | |
| SPO2 peripheral (%) | <0.001*** | |||
| <95 | 524 (8.5) | 404 (7.3) | 120 (18.0) | |
| 95–100 | 2,160 (35.0) | 1,772 (32.0) | 388 (58.0) | |
| Missing | 3,505 (57.0) | 3,340 (61.0) | 165 (25.0) | |
| HR (bpm) | <0.001*** | |||
| <60 | 115 (1.9) | 91 (1.6) | 24 (3.6) | |
| >100 | 525 (8.5) | 393 (7.1) | 132 (20.0) | |
| 60–100 | 2,061 (33.0) | 1,691 (31.0) | 370 (55.0) | |
| Missing | 3,488 (56.0) | 3,341 (61.0) | 147 (22.0) | |
| Warfarin | 926 (15.0) | 882 (16.0) | 44 (6.5) | <0.001*** |
| APSIII | <0.001*** | |||
| >38 | 1,365 (22.0) | 923 (17.0) | 442 (66.0) | |
| 0–38 | 1,484 (24.0) | 1,253 (23.0) | 231 (34.0) | |
| Missing | 3,340 (54.0) | 3,340 (61.0) | 0 (0.0) | |
| SOFA | <0.001*** | |||
| >4 | 1,256 (20.0) | 903 (16.0) | 353 (52.0) | |
| 0–4 | 1,593 (26.0) | 1,273 (23.0) | 320 (48.0) | |
| Missing | 3,340 (54.0) | 3,340 (61.0) | 0 (0.0) | |
| LODS | <0.001*** | |||
| >4 | 1,159 (19.0) | 777 (14.0) | 382 (57.0) | |
| 0–4 | 1,690 (27.0) | 1,399 (25.0) | 291 (43.0) | |
| Missing | 3,340 (54.0) | 3,340 (61.0) | 0 (0.0) | |
| OASIS | <0.001*** | |||
| >30 | 1,325 (21.0) | 920 (17.0) | 405 (60.0) | |
| 0–30 | 1,524 (25.0) | 1,256 (23.0) | 268 (40.0) | |
| Missing | 3,340 (54.0) | 3,340 (61.0) | 0 (0.0) | |
| SIRS | <0.001*** | |||
| >3 | 418 (6.8) | 313 (5.7) | 105 (16.0) | |
| 0–3 | 2,431 (39.0) | 1,863 (34.0) | 568 (84.0) | |
| Missing | 3,340 (54.0) | 3,340 (61.0) | 0 (0.0) | |
| GV | 0.18 (0.12, 0.27) | 0.18 (0.11, 0.27) | 0.20 (0.14, 0.28) | <0.001*** |
| GV_fac | <0.001*** | |||
| Q1 | 1,548 (25.0) | 1,442 (26.0) | 106 (16.0) | |
| Q2 | 1,547 (25.0) | 1,366 (25.0) | 181 (27.0) | |
| Q3 | 1,547 (25.0) | 1,348 (24.0) | 199 (30.0) | |
| Q4 | 1,547 (25.0) | 1,360 (25.0) | 187 (28.0) |
Data are presented as number (%) or median (Q1, Q3). GV index quartile: Q1, 0.004–0.117; Q2, 0.117–0.180; Q3, 0.180–0.273; Q4, 0.273–1.977. *, P<0.05; **, P<0.01; ***, P<0.001. ABC, absolute basophil count; AEC, absolute eosinophil count; AF, atrial fibrillation; ALC, absolute lymphocyte count; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AMC, absolute monocyte count; ANC, absolute neutrophil count; APSIII, acute physiology and chronic health evaluation III; AST, aspartate aminotransferase; BMI, body mass index; CHD, coronary heart disease; CKMB, creatine kinase myocardial band; DBP, diastolic blood pressure; DM, diabetes mellitus; Glu, glucose; GV, glycemic variability; HF, heart failure; HP, hypertension; HR, heart rate; INR, international normalized ratio; LD, lactate dehydrogenase; LODS, logistic organ dysfunction system; OASIS, oxford acute severity of illness score; PCO2, partial pressure of carbon dioxide; pH, potential of hydrogen; PO2, partial pressure of oxygen; PTT, partial thromboplastin time; RBC, red blood cells; RDW, red cell distribution width; RR, respiratory rate; SBP, systolic blood pressure; SIRS, systemic inflammatory response syndrome; SOFA, sequential organ failure assessment; SPO2, oxygen saturation; TB, total bilirubin; total CO2, total carbon dioxide; UN, urea nitrogen; WBC, white blood cells.
The baseline features of patients, divided into quartiles based on the GV index, are shown in Table 2 [interquartile range (IQR) Q1: 0.004–0.117; Q2: 0.117–0.180; Q3: 0.180–0.273; Q4: 0.273–1.977]. The median GV index values for the four groups were 0.08 (0.06, 0.10), 0.15 (0.13, 0.16), 0.22 (0.20, 0.24), and 0.36 (0.31, 0.45), respectively. Patients in the Q4 group were characterized by older age, higher in-hospital mortality, longer hospital stays, more frequent acid-base imbalances, and higher disease severity scores on admission. Additionally, the prevalence of CHD, DM, and HF was notably elevated in this cohort, whereas HP was less common. They also exhibited elevated levels of CKMB, glucose, ALT, AST, creatinine, ALP, UN, LD, WBC, AMC, RDW, and lactate, alongside reduced levels of albumin, RBC, hematocrit, hemoglobin, and SpO2. Furthermore, as GV increased, the proportion of white patients was reduced. AF was found to be significantly more prevalent in the Q4 group when compared to the Q1 group, with all P values under 0.05.
Table 2. Patient characteristics (grouped by IQR of GV).
| Characteristic | Overall (n=6,189) | Q1 (n=1,548) | Q2 (n=1,547) | Q3 (n=1,547) | Q4 (n=1,547) | P value |
|---|---|---|---|---|---|---|
| In-hospital deaths | 176 (2.8) | 19 (1.2) | 32 (2.1) | 48 (3.1) | 77 (5.0) | <0.001*** |
| Length of hospital stay (days) | 6.5 (4.3, 9.8) | 5.1 (3.8, 7.4) | 6.7 (4.6, 9.8) | 7.0 (4.7, 10.9) | 7.4 (4.8, 11.8) | <0.001*** |
| AF | 673 (11.0) | 106 (6.8) | 181 (12.0) | 199 (13.0) | 187 (12.0) | <0.001*** |
| Follow-up time for AF (days) | 5.8 (3.8, 8.8) | 4.9 (3.5, 6.9) | 5.9 (3.8, 8.9) | 6.0 (3.8, 9.5) | 6.6 (4.0, 10.0) | <0.001*** |
| Gender | 0.04* | |||||
| Female | 2,216 (36.0) | 535 (35.0) | 520 (34.0) | 582 (38.0) | 579 (37.0) | |
| Male | 3,973 (64.0) | 1,013 (65.0) | 1,027 (66.0) | 965 (62.0) | 968 (63.0) | |
| Anchor_age (years) | 64 (55, 72) | 63 (54, 71) | 64 (54, 72) | 64 (55, 73) | 65 (56, 73) | 0.01* |
| Marital_status | 0.09 | |||||
| Divorced | 526 (8.5) | 131 (8.5) | 130 (8.4) | 135 (8.7) | 130 (8.4) | |
| Married | 3,267 (53.0) | 809 (52.0) | 838 (54.0) | 802 (52.0) | 818 (53.0) | |
| Null | 224 (3.6) | 41 (2.6) | 54 (3.5) | 74 (4.8) | 55 (3.6) | |
| Single | 1,616 (26.0) | 444 (29.0) | 380 (25.0) | 394 (25.0) | 398 (26.0) | |
| Widowed | 556 (9.0) | 123 (7.9) | 145 (9.4) | 142 (9.2) | 146 (9.4) | |
| Race | <0.001*** | |||||
| Asian | 90 (1.5) | 22 (1.4) | 23 (1.5) | 22 (1.4) | 23 (1.5) | |
| Black | 737 (12.0) | 182 (12.0) | 155 (10.0) | 188 (12.0) | 212 (14.0) | |
| Hispanic | 237 (3.8) | 53 (3.4) | 46 (3.0) | 55 (3.6) | 83 (5.4) | |
| Other | 673 (11.0) | 131 (8.5) | 158 (10.0) | 211 (14.0) | 173 (11.0) | |
| White | 4,452 (72.0) | 1,160 (75.0) | 1,165 (75.0) | 1,071 (69.0) | 1,056 (68.0) | |
| BMI (kg/m2) | 0.07 | |||||
| <18.5 | 39 (0.6) | 10 (0.6) | 9 (0.6) | 10 (0.6) | 10 (0.6) | |
| ≥30 | 3,110 (50.0) | 829 (54.0) | 752 (49.0) | 766 (50.0) | 763 (49.0) | |
| 18.5–24.9 | 463 (7.5) | 113 (7.3) | 116 (7.5) | 133 (8.6) | 101 (6.5) | |
| 25.0–29.9 | 1,028 (17.0) | 254 (16.0) | 278 (18.0) | 239 (15.0) | 257 (17.0) | |
| Missing | 1,549 (25.0) | 342 (22.0) | 392 (25.0) | 399 (26.0) | 416 (27.0) | |
| SBP (mmHg) | 130 (122, 140) | 130 (120, 138) | 130 (122, 140) | 130 (122, 140) | 131 (122, 140) | <0.001*** |
| DBP (mmHg) | 75 (69, 80) | 76 (70, 80) | 75 (70, 80) | 74 (69, 80) | 74 (68, 80) | <0.001*** |
| HP | 2,937 (47.0) | 805 (52.0) | 767 (50.0) | 728 (47.0) | 637 (41.0) | <0.001*** |
| CHD | 2,144 (35.0) | 464 (30.0) | 492 (32.0) | 551 (36.0) | 637 (41.0) | <0.001*** |
| DM | 2,740 (44.0) | 425 (27.0) | 504 (33.0) | 727 (47.0) | 1,084 (70.0) | <0.001*** |
| HF | 2,042 (33.0) | 371 (24.0) | 440 (28.0) | 539 (35.0) | 692 (45.0) | <0.001*** |
| CKMB (ng/mL) | <0.001*** | |||||
| >10 | 290 (4.7) | 52 (3.4) | 75 (4.8) | 82 (5.3) | 81 (5.2) | |
| 0–10 | 1,526 (25.0) | 294 (19.0) | 371 (24.0) | 397 (26.0) | 464 (30.0) | |
| Missing | 4,373 (71.0) | 1,202 (78.0) | 1,101 (71.0) | 1,068 (69.0) | 1,002 (65.0) | |
| INR | 1.23 (1.10, 1.50) | 1.23 (1.10, 1.48) | 1.21 (1.10, 1.50) | 1.24 (1.10, 1.50) | 1.23 (1.10, 1.53) | 0.60 |
| PTT (sec) | 32 (28, 37) | 32 (28, 36) | 32 (28, 36) | 32 (28, 37) | 32 (28, 37) | 0.50 |
| Glu (mg/dL) | 127 (104, 162) | 112 (98, 131) | 124 (105, 146) | 133 (106, 164) | 163 (114, 227) | <0.001*** |
| Albumin (4.2 g/dL) | <0.001*** | |||||
| <3.5 | 1,165 (19.0) | 173 (11.0) | 289 (19.0) | 330 (21.0) | 373 (24.0) | |
| >5.2 | 2 (<0.1) | 0 (0.0) | 1 (<0.1) | 1 (<0.1) | 0 (0.0) | |
| 3.5–5.2 | 1,246 (20.0) | 303 (20.0) | 330 (21.0) | 305 (20.0) | 308 (20.0) | |
| Missing | 3,776 (61.0) | 1,072 (69.0) | 927 (60.0) | 911 (59.0) | 866 (56.0) | |
| ALT (U/L) | <0.001*** | |||||
| >40 | 849 (14.0) | 184 (12.0) | 209 (14.0) | 223 (14.0) | 233 (15.0) | |
| 0–40 | 2,299 (37.0) | 483 (31.0) | 600 (39.0) | 604 (39.0) | 612 (40.0) | |
| Missing | 3,041 (49.0) | 881 (57.0) | 738 (48.0) | 720 (47.0) | 702 (45.0) | |
| AST (U/L) | <0.001*** | |||||
| >40 | 989 (16.0) | 189 (12.0) | 243 (16.0) | 284 (18.0) | 273 (18.0) | |
| 0–40 | 2,173 (35.0) | 484 (31.0) | 568 (37.0) | 543 (35.0) | 578 (37.0) | |
| Missing | 3,027 (49.0) | 875 (57.0) | 736 (48.0) | 720 (47.0) | 696 (45.0) | |
| Creatinine (mg/dL) | 1.00 (0.80, 1.40) | 0.90 (0.80, 1.20) | 1.00 (0.80, 1.30) | 1.00 (0.80, 1.40) | 1.10 (0.90, 1.70) | <0.001*** |
| ALP (U/L) | <0.001*** | |||||
| <40 | 113 (1.8) | 28 (1.8) | 32 (2.1) | 28 (1.8) | 25 (1.6) | |
| >130 | 542 (8.8) | 96 (6.2) | 143 (9.2) | 135 (8.7) | 168 (11.0) | |
| 40–130 | 2,460 (40.0) | 536 (35.0) | 624 (40.0) | 654 (42.0) | 646 (42.0) | |
| Missing | 3,074 (50.0) | 888 (57.0) | 748 (48.0) | 730 (47.0) | 708 (46.0) | |
| UN (mg/dL) | 19 (14, 28) | 17 (13, 22) | 18 (13, 25) | 19 (14, 29) | 22 (15, 37) | <0.001*** |
| LD (U/L) | <0.001*** | |||||
| <94 | 4 (<0.1) | 2 (0.1) | 0 (0.0) | 1 (<0.1) | 1 (<0.1) | |
| >250 | 906 (15.0) | 144 (9.3) | 226 (15.0) | 253 (16.0) | 283 (18.0) | |
| 94–250 | 1,284 (21.0) | 323 (21.0) | 324 (21.0) | 326 (21.0) | 311 (20.0) | |
| Missing | 3,995 (65.0) | 1,079 (70.0) | 997 (64.0) | 967 (63.0) | 952 (62.0) | |
| TB (mg/dL) | <0.001*** | |||||
| >1.5 | 407 (6.6) | 74 (4.8) | 111 (7.2) | 131 (8.5) | 91 (5.9) | |
| 0–1.5 | 2,717 (44.0) | 589 (38.0) | 687 (44.0) | 688 (44.0) | 753 (49.0) | |
| Missing | 3,065 (50.0) | 885 (57.0) | 749 (48.0) | 728 (47.0) | 703 (45.0) | |
| WBC (K/μL) | 9.4 (7.0, 12.4) | 8.9 (6.8, 11.4) | 9.4 (7.0, 12.6) | 9.7 (7.0, 13.0) | 9.5 (7.1, 12.9) | <0.001*** |
| AMC (K/μL) | <0.001*** | |||||
| <0.2 | 111 (1.8) | 15 (1.0) | 29 (1.9) | 29 (1.9) | 38 (2.5) | |
| >0.8 | 506 (8.2) | 94 (6.1) | 138 (8.9) | 137 (8.9) | 137 (8.9) | |
| 0.2–0.8 | 735 (12.0) | 171 (11.0) | 199 (13.0) | 197 (13.0) | 168 (11.0) | |
| Missing | 4,837 (78.0) | 1,268 (82.0) | 1,181 (76.0) | 1,184 (77.0) | 1,204 (78.0) | |
| RDW (%) | 14.20 (13.40, 15.50) | 14.00 (13.20, 15.00) | 14.11 (13.30, 15.30) | 14.21 (13.40, 15.60) | 14.47 (13.50, 15.80) | <0.001*** |
| RBC (m/μL) | 3.95 (3.43, 4.39) | 4.03 (3.51, 4.45) | 3.94 (3.46, 4.38) | 3.94 (3.39, 4.35) | 3.89 (3.37, 4.38) | <0.001*** |
| Hematocrit (%) | 35.6 (31.1, 39.3) | 36.3 (31.9, 39.8) | 35.8 (31.3, 39.3) | 35.3 (30.8, 39.2) | 34.9 (30.6, 39.0) | <0.001*** |
| ALC (K/μL) | <0.001*** | |||||
| <1.2 | 524 (8.5) | 88 (5.7) | 148 (9.6) | 140 (9.0) | 148 (9.6) | |
| >3.7 | 55 (0.9) | 12 (0.8) | 15 (1.0) | 18 (1.2) | 10 (0.6) | |
| 1.2–3.7 | 773 (12.0) | 180 (12.0) | 203 (13.0) | 205 (13.0) | 185 (12.0) | |
| Missing | 4,837 (78.0) | 1,268 (82.0) | 1,181 (76.0) | 1,184 (77.0) | 1,204 (78.0) | |
| ABC (K/μL) | <0.001*** | |||||
| <0.01 | 175 (2.8) | 19 (1.2) | 43 (2.8) | 58 (3.7) | 55 (3.6) | |
| >0.08 | 73 (1.2) | 10 (0.6) | 17 (1.1) | 21 (1.4) | 25 (1.6) | |
| 0.01–0.08 | 1,104 (18.0) | 251 (16.0) | 306 (20.0) | 284 (18.0) | 263 (17.0) | |
| Missing | 4,837 (78.0) | 1,268 (82.0) | 1,181 (76.0) | 1,184 (77.0) | 1,204 (78.0) | |
| AEC (K/μL) | <0.001*** | |||||
| <0.04 | 363 (5.9) | 53 (3.4) | 99 (6.4) | 107 (6.9) | 104 (6.7) | |
| >0.54 | 51 (0.8) | 8 (0.5) | 16 (1.0) | 15 (1.0) | 12 (0.8) | |
| 0.04–0.54 | 938 (15.0) | 219 (14.0) | 251 (16.0) | 241 (16.0) | 227 (15.0) | |
| Missing | 4,837 (78.0) | 1,268 (82.0) | 1,181 (76.0) | 1,184 (77.0) | 1,204 (78.0) | |
| Hemoglobin (g/dL) | 11.60 (10.10, 13.00) | 11.90 (10.40, 13.20) | 11.70 (10.20, 13.00) | 11.50 (10.00, 12.90) | 11.30 (9.80, 12.90) | <0.001*** |
| Platelets (K/μL) | 211 (162, 260) | 211 (163, 256) | 209 (159, 257) | 211 (161, 261) | 212 (167, 266) | 0.10 |
| ANC (K/μL) | 32 (28, 37) | 32 (28, 37) | 32 (28, 36) | 32 (28, 37) | 32 (28, 38) | 0.40 |
| pH | <0.001*** | |||||
| <7.35 | 907 (15.0) | 126 (8.1) | 232 (15.0) | 254 (16.0) | 295 (19.0) | |
| >7.45 | 326 (5.3) | 48 (3.1) | 94 (6.1) | 78 (5.0) | 106 (6.9) | |
| 7.35–7.45 | 1,705 (28.0) | 332 (21.0) | 434 (28.0) | 478 (31.0) | 461 (30.0) | |
| Missing | 3,251 (53.0) | 1,042 (67.0) | 787 (51.0) | 737 (48.0) | 685 (44.0) | |
| Total CO2 (mEq/L) | <0.001*** | |||||
| <21 | 219 (3.5) | 25 (1.6) | 43 (2.8) | 68 (4.4) | 83 (5.4) | |
| >30 | 623 (10.0) | 104 (6.7) | 141 (9.1) | 157 (10.0) | 221 (14.0) | |
| 21–30 | 2,006 (32.0) | 362 (23.0) | 555 (36.0) | 558 (36.0) | 531 (34.0) | |
| Missing | 3,341 (54.0) | 1,057 (68.0) | 808 (52.0) | 764 (49.0) | 712 (46.0) | |
| Lactate (mmol/L) | <0.001*** | |||||
| <0.5 | 6 (<0.1) | 1 (<0.1) | 2 (0.1) | 1 (<0.1) | 2 (0.1) | |
| >2 | 681 (11.0) | 82 (5.3) | 166 (11.0) | 199 (13.0) | 234 (15.0) | |
| 0.5–2 | 2,137 (35.0) | 409 (26.0) | 582 (38.0) | 567 (37.0) | 579 (37.0) | |
| Missing | 3,365 (54.0) | 1,056 (68.0) | 797 (52.0) | 780 (50.0) | 732 (47.0) | |
| PCO2 (mmHg) | <0.001*** | |||||
| <35 | 339 (5.5) | 48 (3.1) | 90 (5.8) | 96 (6.2) | 105 (6.8) | |
| >45 | 1,203 (19.0) | 206 (13.0) | 293 (19.0) | 330 (21.0) | 374 (24.0) | |
| 35–45 | 1,304 (21.0) | 236 (15.0) | 356 (23.0) | 356 (23.0) | 356 (23.0) | |
| Missing | 3,343 (54.0) | 1,058 (68.0) | 808 (52.0) | 765 (49.0) | 712 (46.0) | |
| PO2 (mmHg) | <0.001*** | |||||
| <85 | 1,004 (16.0) | 152 (9.8) | 222 (14.0) | 290 (19.0) | 340 (22.0) | |
| >105 | 1,544 (25.0) | 282 (18.0) | 440 (28.0) | 405 (26.0) | 417 (27.0) | |
| 85–105 | 298 (4.8) | 56 (3.6) | 77 (5.0) | 87 (5.6) | 78 (5.0) | |
| Missing | 3,343 (54.0) | 1,058 (68.0) | 808 (52.0) | 765 (49.0) | 712 (46.0) | |
| RR (bpm) | <0.001*** | |||||
| <12 | 236 (3.8) | 53 (3.4) | 56 (3.6) | 67 (4.3) | 60 (3.9) | |
| >20 | 785 (13.0) | 123 (7.9) | 215 (14.0) | 219 (14.0) | 228 (15.0) | |
| 12–20 | 1,681 (27.0) | 354 (23.0) | 455 (29.0) | 440 (28.0) | 432 (28.0) | |
| Missing | 3,487 (56.0) | 1,018 (66.0) | 821 (53.0) | 821 (53.0) | 827 (53.0) | |
| SPO2 peripheral (%) | <0.001*** | |||||
| <95 | 524 (8.5) | 84 (5.4) | 134 (8.7) | 152 (9.8) | 154 (10.0) | |
| 95–100 | 2,160 (35.0) | 440 (28.0) | 587 (38.0) | 573 (37.0) | 560 (36.0) | |
| Missing | 3,505 (57.0) | 1,024 (66.0) | 826 (53.0) | 822 (53.0) | 833 (54.0) | |
| HR (bpm) | <0.001*** | |||||
| <60 | 115 (1.9) | 34 (2.2) | 28 (1.8) | 30 (1.9) | 23 (1.5) | |
| >100 | 525 (8.5) | 75 (4.8) | 142 (9.2) | 145 (9.4) | 163 (11.0) | |
| 60–100 | 2,061 (33.0) | 421 (27.0) | 555 (36.0) | 551 (36.0) | 534 (35.0) | |
| Missing | 3,488 (56.0) | 1,018 (66.0) | 822 (53.0) | 821 (53.0) | 827 (53.0) | |
| Warfarin | 926 (15.0) | 211 (14.0) | 234 (15.0) | 236 (15.0) | 245 (16.0) | 0.40 |
| APSIII | <0.001*** | |||||
| >38 | 1,365 (22.0) | 177 (11.0) | 323 (21.0) | 398 (26.0) | 467 (30.0) | |
| 0–38 | 1,484 (24.0) | 384 (25.0) | 435 (28.0) | 368 (24.0) | 297 (19.0) | |
| Missing | 3,340 (54.0) | 987 (64.0) | 789 (51.0) | 781 (50.0) | 783 (51.0) | |
| SOFA | <0.001*** | |||||
| >4 | 1,256 (20.0) | 183 (12.0) | 316 (20.0) | 364 (24.0) | 393 (25.0) | |
| 0–4 | 1,593 (26.0) | 378 (24.0) | 442 (29.0) | 402 (26.0) | 371 (24.0) | |
| Missing | 3,340 (54.0) | 987 (64.0) | 789 (51.0) | 781 (50.0) | 783 (51.0) | |
| LODS | <0.001*** | |||||
| >4 | 1,159 (19.0) | 140 (9.0) | 279 (18.0) | 347 (22.0) | 393 (25.0) | |
| 0–4 | 1,690 (27.0) | 421 (27.0) | 479 (31.0) | 419 (27.0) | 371 (24.0) | |
| Missing | 3,340 (54.0) | 987 (64.0) | 789 (51.0) | 781 (50.0) | 783 (51.0) | |
| OASIS | <0.001*** | |||||
| >30 | 1,325 (21.0) | 202 (13.0) | 339 (22.0) | 381 (25.0) | 403 (26.0) | |
| 0–30 | 1,524 (25.0) | 359 (23.0) | 419 (27.0) | 385 (25.0) | 361 (23.0) | |
| Missing | 3,340 (54.0) | 987 (64.0) | 789 (51.0) | 781 (50.0) | 783 (51.0) | |
| SIRS | <0.001*** | |||||
| >3 | 418 (6.8) | 68 (4.4) | 95 (6.1) | 122 (7.9) | 133 (8.6) | |
| 0–3 | 2,431 (39.0) | 493 (32.0) | 663 (43.0) | 644 (42.0) | 631 (41.0) | |
| Missing | 3,340 (54.0) | 987 (64.0) | 789 (51.0) | 781 (50.0) | 783 (51.0) | |
| GV | 0.18 (0.12, 0.27) | 0.08 (0.06, 0.10) | 0.15 (0.13, 0.16) | 0.22 (0.20, 0.24) | 0.36 (0.31, 0.45) | <0.001*** |
Data are presented as number (%) or median (Q1, Q3). GV index quartile: Q1, 0.004–0.117; Q2, 0.117–0.180; Q3, 0.180–0.273; Q4, 0.273–1.977. *, P<0.05; ***, P<0.001. ABC, absolute basophil count; AEC, absolute eosinophil count; AF, atrial fibrillation; ALC, absolute lymphocyte count; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AMC, absolute monocyte count; ANC, absolute neutrophil count; APSIII, acute physiology and chronic health evaluation III; AST, aspartate aminotransferase; BMI, body mass index; CHD, coronary heart disease; CKMB, creatine kinase myocardial band; DBP, diastolic blood pressure; DM, diabetes mellitus; Glu, glucose; GV, glycemic variability; HF, heart failure; HP, hypertension; HR, heart rate; INR, international normalized ratio; IQR, interquartile range; LD, lactate dehydrogenase; LODS, logistic organ dysfunction system; OASIS, oxford acute severity of illness score; pH, potential of hydrogen; PO2, partial pressure of oxygen; PTT, partial thromboplastin time; RBC, red blood cell; RDW, red cell distribution width; RR, respiratory rate; SBP, systolic blood pressure; SIRS, systemic inflammatory response syndrome; SOFA, sequential organ failure assessment; TB, total bilirubin; total CO2, total carbon dioxide; UN, urea nitrogen; WBC, white blood cell.
Primary outcomes
The LR analysis revealed a notable relationship between the GV index, considered as a continuous variable, and the risk of developing AF in the unadjusted (OR 2.579, 95% CI: 1.595–4.103; P<0.001), partially adjusted (OR 2.736, 95% CI: 1.659–4.447; P<0.001), and fully adjusted (OR 2.003, 95% CI: 1.052–3.749; P=0.03) models. Similarly, in the unadjusted (OR 5.905, 95% CI: 2.848–11.606; P<0.001) and partially adjusted (OR 5.987, 95% CI: 2.830–12.013; P<0.001) models, in-hospital mortality was significantly correlated with GV index. However, the fully adjusted model revealed no notable correlation of in-hospital mortality with GV index.
Additionally, when stratifying the GV index into quartiles, the fourth quartile (Q4) revealed a strong link to the risk of developing AF across all models: unadjusted (OR 1.871, 95% CI: 1.461–2.407; P<0.001), partially adjusted (OR 1.858, 95% CI: 1.445–2.401; P<0.001), and fully adjusted (OR 1.451, 95% CI: 1.082–1.951; P=0.01). Similarly, there was a strong correlation between the Q4 GV index and the risk of mortality during hospitalization in the unadjusted (OR, 4.215, 95% CI: 2.595–7.198; P<0.001), partially adjusted (OR 4.107, 95% CI: 2.524–7.026; P<0.001), and fully adjusted models (OR 1.996, 95% CI: 1.137–3.640; P=0.02) (Table 3).
Table 3. The relationship between GV and the risks of AF and hospitalization-related death.
| Characteristic | ARD (95% CI) | Model 1 | Model 2 | Model 3 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P value | P of trend | OR (95% CI) | P value | P of trend | OR (95% CI) | P value | P of trend | ||||
| AF incidence | ||||||||||||
| GV (continuous) | 2.579 (1.595–4.103) |
<0.001 | 2.736 (1.659–4.447) |
<0.001 | 2.003 (1.052–3.749) |
0.03 | ||||||
| GV (IQR) | <0.001 | <0.001 | 0.02 | |||||||||
| Q1 | −0.054 (−0.069 to −0.038) |
– | – | – | ||||||||
| Q2 | 0.011 (−0.007 to 0.029) |
1.803 (1.405–2.322) |
<0.001 | 1.775 (1.378–2.296) |
<0.001 | 1.599 (1.214–2.115) |
<0.001 | |||||
| Q3 | 0.027 (0.008 to 0.045) |
2.008 (1.572–2.578) |
<0.001 | 1.935 (1.508–2.496) |
<0.001 | 1.619 (1.226–2.146) |
<0.001 | |||||
| Q4 | 0.016 (−0.002 to 0.035) |
1.871 (1.461–2.407) |
<0.001 | 1.858 (1.445–2.401) |
<0.001 | 1.451 (1.082–1.951) |
0.01 | |||||
| In-hospital mortality rate | ||||||||||||
| GV (continuous) | 5.905 (2.848–11.606) |
<0.001 | 5.987 (2.830–12.013) |
<0.001 | 2.46 (0.835–6.669) |
0.09 | ||||||
| GV (IQR) | <0.001 | <0.001 | 0.02 | |||||||||
| Q1 | −0.022 (−0.029 to −0.014) |
– | – | – | – | – | ||||||
| Q2 | −0.010 (−0.019 to −0.002) |
1.7 (0.969–3.065) |
0.07 | 1.646 (0.937–2.971) |
0.089 | 1.036 (0.565–1.944) |
0.91 | |||||
| Q3 | 0.003 (−0.006 to 0.013) |
2.577 (1.533–4.509) |
<0.001 | 2.409 (1.430–4.224) |
0.001 | 1.235 (0.695–2.266) |
0.48 | |||||
| Q4 | 0.028 (0.017 to 0.040) |
4.215 (2.595–7.198) |
<0.001 | 4.107 (2.524–7.026) |
<0.001 | 1.996 (1.137–3.640) |
0.02 | |||||
GV index quartile: Q1, 0.004–0.117; Q2, 0.117–0.180; Q3, 0.180–0.273; Q4, 0.273–1.977. Model 1 was unadjusted. Model 2 was adjusted for age, gender and race. Model 3 was adjusted for age, gender, race, HP, DM, HF, BMI, CHD, CKMB, INR, warfarin, albumin, ALT, AST, creatinine, ALP, UN, TB, WBC, AMC, RDW, RBC, hematocrit, hemoglobin, platelets, pH, PO2 and HR. AF, atrial fibrillation; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AMC, absolute monocyte count; ARD, absolute risk differences; AST, aspartate aminotransferase; BMI, body mass index; CHD, coronary heart disease; CI, confidence interval; CKMB, creatine kinase myocardial band; DM, diabetes mellitus; GV, glycemic variability; HF, heart failure; HP, hypertension; HR, heart rate; INR, international normalized ratio; IQR, interquartile range; OR, odds ratio; pH, potential of hydrogen; PO2, partial pressure of oxygen; RBC, red blood cell; RDW, red blood cell distribution width; TB, total bilirubin; UN, urea nitrogen; WBC, white blood cell.
The cumulative incidence curve illustrated in Figure 2A reveals a significant variation in the occurrence of AF among patients categorized by different quartiles of GV (P<0.001). The survival curve illustrated in Figure 2B, categorized by quartiles of the GV index, reveals a notable statistical difference in survival rates among patients exhibiting varying levels of GV (P=0.009).
Figure 2.
The cumulative event incidence curves for incidence of AF (A) and hospitalized patients’ survival probability curve (B). GV index quartile: Q1, 0.004–0.117; Q2, 0.117–0.180; Q3, 0.180–0.273; Q4, 0.273–1.977. AF, atrial fibrillation; GV, glycemic variability.
The RCS illustrated in Figure 3 reveals a correlation of GV index with AF as well as in-hospital mortality, indicating a dose-response relationship in the unadjusted, partially adjusted, and fully adjusted models (nonlinear P<0.001, P<0.001, P<0.001, P<0.001, P<0.001, and P=0.04).
Figure 3.
Restricted cubic spline curves for the GV index hazard ratio. (A,D) Model 1 was unadjusted. (B,E) Model 2 was adjusted for age, gender and race. (C,F) Model 3 was adjusted for age, gender, race, HP, DM, HF, BMI, CHD, CKMB, INR, warfarin, albumin, ALT, AST, creatinine, ALP, UN, TB, WBC, AMC, RDW, RBC, hematocrit, hemoglobin, platelets, pH, PO2 and HR. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AMC, absolute monocyte count; AST, aspartate aminotransferase; BMI, body mass index; CHD, coronary heart disease; CI, confidence interval; CKMB, creatine kinase myocardial band; DM, diabetes mellitus; GV, glycemic variability; HF, heart failure; HP, hypertension; HR, heart rate; INR, international normalized ratio; OR, odds ratio; pH, potential of hydrogen; PO2, partial pressure of oxygen; RBC, red blood cell; RCS, restricted cubic spline; RDW, red blood cell distribution width; TB, total bilirubin; UN, urea nitrogen; WBC, white blood cell.
Subsequently, we performed subgroup analyses by age, sex, race, BMI, DM, HP, and HF (Figure 4). The findings indicated that GV was significantly correlated with AF risk across several subgroups: males (OR 2.906, 95% CI: 1.667–5.068; P<0.001), those aged over 60 years (OR 3.02, 95% CI: 1.721–5.301; P<0.001), White patients (OR 3.077, 95% CI: 1.693–5.594; P<0.001), individuals whose BMI was ≥30 kg/m2 (OR 4.999, 95% CI: 1.778–14.053; P=0.002) or between 25.0 and 29.9 kg/m2 (OR 2.247, 95% CI: 1.021–4.946; P=0.04), patients with DM (OR 2.355, 95% CI: 1.146–4.836; P=0.02) and without DM (OR 3.213, 95% CI: 1.666–6.196; P<0.001), those with HP (OR 2.583, 95% CI: 1.226–5.445; P=0.01) and without HP (OR 2.437, 95% CI: 1.32–4.501; P=0.004), and those without HF (OR 2.36, 95% CI: 1.175–4.739; P=0.02). There was a significant correlation between GV and in-hospital mortality across the following subgroups: females (OR 8.653, 95% CI: 2.583–28.992; P<0.001), males (OR 4.977, 95% CI: 2.115–11.712; P<0.001), individuals aged ≤60 years (OR 7.51, 95% CI: 2.498–22.572; P<0.001) and >60 years (OR 5.128, 95% CI: 2.072–12.691; P<0.001), White patients (OR 8.829, 95% CI: 3.479–22.406; P<0.001), those with a BMI of 25.0–29.9 kg/m2 (OR 12.272, 95% CI: 2.964–50.808; P=0.001), patients without DM (OR 4.85, 95% CI: 1.899–12.385; P=0.001) and with DM (OR 9.555, 95% CI: 3.129–29.182; P<0.001), those without HP (OR 4.286, 95% CI: 1.757–10.459; P=0.001) and with HP (OR 8.409, 95% CI: 2.666–26.526; P<0.001), those without HF (OR 7.074, 95% CI: 2.772–18.053; P<0.001) and with HF (OR 3.155, 95% CI 1.03–9.662; P=0.04). In the subgroup evaluations, no significant interaction was found between the variables and the GV index, as all P values for interaction were greater than 0.05.
Figure 4.
Forest plots of hazard ratios for the primary endpoint in different subgroups. (A) AF risk. (B) In-hospital mortality rate. AF, atrial fibrillation; BMI, body mass index; CI, confidence interval; DM, diabetes Mellitus; HF, heart failure; HP, hypertension; OR, odds ratio.
Discussion
To our knowledge, this research is the first retrospective study to analyze the correlation of the GV index with AF and in-hospital mortality in OSA patients, presenting a new angle for comprehending the mechanisms associated with cardiovascular risk. Our research indicates that a higher GV index is closely linked to the development of AF and increased mortality rates during hospitalization among OSA patients. However, after adjustment for multiple confounding factors, it is noteworthy that the strength of the association between GV and different clinical outcomes was not consistent. While the link between GV and AF remained robust in the multifactorial analysis, its association with in-hospital mortality was attenuated. This finding suggests that the impact of GV on AF may be more direct and specific, whereas its relationship with mortality is likely subject to greater interference from confounding factors or is mediated indirectly through the induction of other complications, such as AF. Consequently, while the independent effect of GV on mortality should not be overinterpreted, its clear role in the pathogenesis of AF warrants full attention.
Sleep fragmentation and intermittent hypoxia in OSA patients can heighten sympathetic nervous system activity, which may lead to electrophysiological instability of the atria, thereby predisposing individuals to AF (16,17). Furthermore, OSA is frequently associated with metabolic disorders such as insulin resistance and obesity (18), both of which are linked to altered atrial electrophysiology (19). A rise in metabolic issues could worsen GV, and significant OSA frequently correlates with elevated levels of GV (20). Elevated GV indicates greater fluctuations in blood glucose levels, which can impair cardiac adaptation to metabolic changes, lead to cardiac fibrosis, and heighten the risk of experiencing AF (21). Moreover, pronounced glycemic fluctuations can significantly raise oxidative stress in endothelial cells, enhance monocyte-endothelial adhesion, heighten the rate of endothelial cell death and dysfunction, and result in more serious cardiovascular harm (22), thereby further contributing to AF development. Our findings demonstrate that in OSA patients, AF incidence rises significantly with increased GV, and this elevation is closely linked to higher levels of pro-inflammatory markers, such as WBC counts. Sleep fragmentation and intermittent hypoxia often result in hypoxemia during sleep in OSA patients, which can provoke a chronic inflammatory response (23). Increased levels of GV could enhance the binding of inflammatory cytokines to the cells lining blood vessels, thereby amplifying the overall inflammatory response in vivo (24). Structural and functional abnormalities of the atria, driven by chronic inflammation (25), have been shown to contribute to the onset of AF (26).
OSA is frequently accompanied by several comorbidities, including metabolic syndrome, hypertension, obesity, and DM (27). It can lead to various harmful consequences, including disrupted sleep patterns, varying pressures within the chest cavity, temporary oxygen deprivation, changes in the activity of the sympathetic nervous system, heightened systemic inflammation, and other unexplained risk factors or OSA-related mechanisms (28). When combined, these variables can influence multi-organ failure or systemic issues, either directly or indirectly, thereby increasing the risk of mortality in patients. New findings reveal that a significant risk of overall death is linked to severe OSA (29). Disrupted sleep patterns are viewed as a significant factor leading to negative effects associated with OSA, and the frequency of awakenings has been associated with a higher risk of mortality from all causes in older individuals (30). High GV not only correlates with the occurrence of atrial AF, but also with other cardiovascular conditions such as CHD, HF, and chronic inflammation. A comprehensive study involving 7,049 patients in critical condition who received regular glucose assessments revealed that GV serves as a standalone indicator of death rates in the ICU as well as overall in-hospital mortality (31). Research conducted by Van den Berghe et al. revealed that rigorous insulin management in the ICU was associated with lower mortality rates (32), though a subsequent large trial did not confirm these findings (33). The recent research, which utilized logistic regression techniques, found that hypoglycemia serves as a standalone risk factor for increased mortality rates (33). Lower blood glucose levels might have contributed to greater glucose fluctuations, which were not directly assessed in these studies but might be essential. Current investigations have been paying more attention to the value of GV. According to Mo et al., elevated short-term GV correlates with a heightened risk of mortality from all causes, even among patients with well-managed T2DM who are using continuous glucose monitoring (34). Similarly, a study by Chun and associates indicated that significant in-hospital variations in glucose levels before patients were discharged were tied to a rise in overall mortality within the following year, notably in non-diabetic patients with acute HF (35). According to a meta-analysis by Lin et al., the presence of acute GV is identified as a significant independent risk factor contributing to early death in individuals suffering from acute strokes (36). A separate prospective analysis found that elevated levels of GV were independently related to higher mortality among coronavirus disease 2019 (COVID-19) patients, regardless of the presence of DM (37). The evidence gathered indicates a strong link between GV and mortality, aligning with our own findings. This evidence highlights the potential importance of GV management in improving patient outcomes across diverse populations, not limited to any specific group.
Prior research has established GV as a significant independent predictor of mortality and morbidity in critically ill patients admitted to the ICU (38,39). Similarly, in populations with DM, GV is well-recognized risk factors for diabetic complications and cardiovascular events (40,41). While our findings in OSA patients resonate with this established theme—that GV portends poorer health outcomes—they also highlight a crucial and underappreciated dimension. The novelty of our study lies in identifying GV as a key risk factor within the unique pathophysiological milieu of OSA. Unlike the acute severe stress and iatrogenic factors that often drive GV in the ICU, or the primary insulin resistance and beta-cell dysfunction in diabetes, the mechanisms in OSA are likely distinct. We postulate a self-perpetuating cycle: OSA-induced sleep fragmentation and intermittent hypoxia can directly promote GV through sympathetic activation and hormonal dysregulation (42,43). This elevated GV, in turn, may exacerbate oxidative stress, systemic inflammation (44), and atrial remodeling (45), thereby culminating in AF and potentially contributing to mortality. This pathophysiological triad—linking sleep-disordered breathing, glucose dysregulation, and cardiovascular injury—is far less explored than in other settings. Therefore, our work extends the implications of GV beyond traditional high-risk groups and positions it as a modifiable risk factor at the intersection of sleep medicine and cardiometabolic disease. It suggests that OSA patients, even in the absence of overt diabetes or critical illness, may constitute a population that stands to benefit significantly from GV assessment and management.
In summary, GV, as a multidimensional biomarker, may serve as an indicator of overall health status rather than merely reflecting glucose metabolism stability. Patients with OSA frequently present with multiple comorbidities, which, when combined with disease progression, can lead to more pronounced fluctuations in GV, further compounding the patient’s health burden. The interaction between OSA and elevated GV may give rise to complex pathophysiological changes, warranting further research to explore and deepen our understanding of this intricate relationship. Our study results indicate that in OSA patients, elevated GV is not only significantly associated with the occurrence of AF but also with increased in-hospital mortality. These insights offer new perspectives and strategies for managing patients with OSA, underscoring the critical importance of vigilant GV monitoring. Incorporating GV into cardiovascular health assessments for OSA patients could provide an opportunity for targeted interventions aimed at improving patient outcomes. Future clinical practice should prioritize GV management as a key element in the comprehensive care of OSA patients to enhance prognostic outcomes through tailored approaches.
There are several drawbacks associated with this research. Firstly, due to its retrospective nature, it is incapable of conclusively demonstrating a causal connection between GV and AF. Secondly, the significant presence of Caucasians in the database constrains the ability to generalize our results to various racial and ethnic groups. Thirdly, certain variables of interest, such as dietary habits, educational background, respiratory sleep monitoring, and continuous positive airway pressure (CPAP) therapy, were not included due to the constraints of the available data. Additionally, the exclusion of patients with fewer than three glucose measurements may introduce selection bias toward more critically ill or diabetic patients. Finally, to avoid selection bias caused by sample exclusion, we classified the missing values for variables with a missing rate greater than 25%. However, this omission might also lead to potential missing bias, where the impact of unknown data (missing) on the results is ignored.
Conclusions
A significant correlation exists between a high GV index and an elevated risk of both AF and in-hospital mortality in OSA patients. The study indicates that the GV index is valuable in anticipating both the incidence of AF and in-hospital mortality, potentially facilitating the identification of individuals with a higher risk profile. Furthermore, GV assessment may refine risk stratification and guide personalized glycemic management, ultimately improving patient outcomes. Future research can explore the pathophysiological mechanisms underlying high GV in OSA patients.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.
Footnotes
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1678/rc
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1678/coif). The authors have no conflicts of interest to declare.
References
- 1.Liu L, He G, Yu R, et al. Causal relationships between gut microbiome and obstructive sleep apnea: a bi-directional Mendelian randomization. Front Microbiol 2024;15:1410624. 10.3389/fmicb.2024.1410624 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sánchez-de-la-Torre M, Sánchez-de-la-Torre A, Bertran S, et al. Effect of obstructive sleep apnoea and its treatment with continuous positive airway pressure on the prevalence of cardiovascular events in patients with acute coronary syndrome (ISAACC study): a randomised controlled trial. Lancet Respir Med 2020;8:359-67. 10.1016/S2213-2600(19)30271-1 [DOI] [PubMed] [Google Scholar]
- 3.Maniaci A, Lavalle S, Anzalone R, et al. Oral Health Implications of Obstructive Sleep Apnea: A Literature Review. Biomedicines 2024;12:1382. 10.3390/biomedicines12071382 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pirruccello JP, Di Achille P, Choi SH, et al. Deep learning of left atrial structure and function provides link to atrial fibrillation risk. Nat Commun 2024;15:4304. 10.1038/s41467-024-48229-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lip GY, Fauchier L, Freedman SB, et al. Atrial fibrillation. Nat Rev Dis Primers 2016;2:16016. 10.1038/nrdp.2016.16 [DOI] [PubMed] [Google Scholar]
- 6.Yapanis M, James S, Craig ME, et al. Complications of Diabetes and Metrics of Glycemic Management Derived From Continuous Glucose Monitoring. J Clin Endocrinol Metab 2022;107:e2221-36. 10.1210/clinem/dgac034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yeghiazarians Y, Jneid H, Tietjens JR, et al. Obstructive Sleep Apnea and Cardiovascular Disease: A Scientific Statement From the American Heart Association. Circulation 2021;144:e56-67. 10.1161/CIR.0000000000000988 [DOI] [PubMed] [Google Scholar]
- 8.Redline S, Azarbarzin A, Peker Y. Obstructive sleep apnoea heterogeneity and cardiovascular disease. Nat Rev Cardiol 2023;20:560-73. 10.1038/s41569-023-00846-6 [DOI] [PubMed] [Google Scholar]
- 9.Bouzerda A. Cardiovascular risk and obstructive sleep apnea. Pan Afr Med J 2018;29:47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.May AM, Van Wagoner DR, Mehra R. OSA and Cardiac Arrhythmogenesis: Mechanistic Insights. Chest 2017;151:225-41. 10.1016/j.chest.2016.09.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Saito K, Okada Y, Torimoto K, et al. Blood glucose dynamics during sleep in patients with obstructive sleep apnea and normal glucose tolerance: effects of CPAP therapy. Sleep Breath 2022;26:771-81. 10.1007/s11325-021-02442-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sircu V, Colesnic SI, Covantsev S, et al. The Burden of Comorbidities in Obstructive Sleep Apnea and the Pathophysiologic Mechanisms and Effects of CPAP. Clocks Sleep 2023;5:333-49. 10.3390/clockssleep5020025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tay J, Thompson CH, Brinkworth GD. Glycemic Variability: Assessing Glycemia Differently and the Implications for Dietary Management of Diabetes. Annu Rev Nutr 2015;35:389-424. 10.1146/annurev-nutr-121214-104422 [DOI] [PubMed] [Google Scholar]
- 14.Johnson AEW, Bulgarelli L, Shen L, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data 2023;10:1. 10.1038/s41597-022-01899-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.He HM, Zheng SW, Xie YY, et al. Simultaneous assessment of stress hyperglycemia ratio and glycemic variability to predict mortality in patients with coronary artery disease: a retrospective cohort study from the MIMIC-IV database. Cardiovasc Diabetol 2024;23:61. 10.1186/s12933-024-02146-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Che T, Yan C, Tian D, et al. The Association Between Sleep and Metabolic Syndrome: A Systematic Review and Meta-Analysis. Front Endocrinol (Lausanne) 2021;12:773646. 10.3389/fendo.2021.773646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Tafelmeier M, Kuettner S, Hauck C, et al. Sleep-Disordered Breathing, Advanced Age, and Diabetes Mellitus Are Associated with De Novo Atrial Fibrillation after Cardiac Surgery. Biomedicines 2024;12:1035. 10.3390/biomedicines12051035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Guo Y, Sun S, Wang Y, et al. Microbial dysbiosis in obstructive sleep apnea: a systematic review and meta-analysis. Front Microbiol 2025;16:1572637. 10.3389/fmicb.2025.1572637 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ding M, Wennberg A, Gigante B, et al. Lipid levels in midlife and risk of atrial fibrillation over 3 decades-Experience from the Swedish AMORIS cohort: A cohort study. PLoS Med 2022;19:e1004044. 10.1371/journal.pmed.1004044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Aurora RN, Gaynanova I, Patel P, et al. Glucose profiles in obstructive sleep apnea and type 2 diabetes mellitus. Sleep Med 2022;95:105-11. 10.1016/j.sleep.2022.04.007 [DOI] [PubMed] [Google Scholar]
- 21.Lee HJ, Choi EK, Han KD, et al. High variability in bodyweight is associated with an increased risk of atrial fibrillation in patients with type 2 diabetes mellitus: a nationwide cohort study. Cardiovasc Diabetol 2020;19:78. 10.1186/s12933-020-01059-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Manosroi W, Phimphilai M, Waisayanand N, et al. Glycated hemoglobin variability and the risk of cardiovascular events in patients with prediabetes and type 2 diabetes mellitus: A post-hoc analysis of a prospective and multicenter study. J Diabetes Investig 2023;14:1391-400. 10.1111/jdi.14073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Liu Y, Tan SX, Wu YK, et al. Altered Intrinsic Regional Spontaneous Brain Activity in Patients With Severe Obesity and Meibomian Gland Dysfunction: A Resting-State Functional Magnetic Resonance Imaging Study. Front Hum Neurosci 2022;16:879513. 10.3389/fnhum.2022.879513 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Cai Y, Zhang H, Li Q, et al. Correlation Between Blood Glucose Variability and Early Therapeutic Effects After Intravenous Thrombolysis With Alteplase and Levels of Serum Inflammatory Factors in Patients With Acute Ischemic Stroke. Front Neurol 2022;13:806013. 10.3389/fneur.2022.806013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Fan S, Hu Y. Role of m6A Methylation in the Occurrence and Development of Heart Failure. Front Cardiovasc Med 2022;9:892113. 10.3389/fcvm.2022.892113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yang YH, Chiang HJ, Yip HK, et al. Risk of New-Onset Atrial Fibrillation Among Asian Chronic Hepatitis C Virus Carriers: A Nationwide Population-Based Cohort Study. J Am Heart Assoc 2019;8:e012914. 10.1161/JAHA.119.012914 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Sultana R, Sissoho F, Kaushik VP, et al. The Case for Early Use of Glucagon-like Peptide-1 Receptor Agonists in Obstructive Sleep Apnea Patients with Comorbid Diabetes and Metabolic Syndrome. Life (Basel) 2022;12:1222. 10.3390/life12081222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bushi G, Padhi BK, Shabil M, et al. Cardiovascular Disease Outcomes Associated with Obstructive Sleep Apnea in Diabetics: A Systematic Review and Meta-Analysis. Diseases 2023;11:103. 10.3390/diseases11030103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.US Preventive Services Task Force ; Mangione CM, Barry MJ, et al. Screening for Obstructive Sleep Apnea in Adults: US Preventive Services Task Force Recommendation Statement. JAMA 2022;328:1945-50. 10.1001/jama.2022.20304 [DOI] [PubMed] [Google Scholar]
- 30.Lin Y, Wu Y, Lin Q, et al. Objective Sleep Duration and All-Cause Mortality Among People With Obstructive Sleep Apnea. JAMA Netw Open 2023;6:e2346085. 10.1001/jamanetworkopen.2023.46085 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ouattara A, Grimaldi A, Riou B. Blood glucose variability: a new paradigm in critical care? Anesthesiology 2006;105:233-4. 10.1097/00000542-200608000-00002 [DOI] [PubMed] [Google Scholar]
- 32.van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-67. 10.1056/NEJMoa011300 [DOI] [PubMed] [Google Scholar]
- 33.Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449-61. 10.1056/NEJMoa052521 [DOI] [PubMed] [Google Scholar]
- 34.Mo Y, Wang C, Lu J, et al. Impact of short-term glycemic variability on risk of all-cause mortality in type 2 diabetes patients with well-controlled glucose profile by continuous glucose monitoring: A prospective cohort study. Diabetes Res Clin Pract 2022;189:109940. 10.1016/j.diabres.2022.109940 [DOI] [PubMed] [Google Scholar]
- 35.Chun KH, Oh J, Lee CJ, et al. In-hospital glycemic variability and all-cause mortality among patients hospitalized for acute heart failure. Cardiovasc Diabetol 2022;21:291. 10.1186/s12933-022-01720-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Lin J, Cai C, Xie Y, et al. Acute glycemic variability and mortality of patients with acute stroke: a meta-analysis. Diabetol Metab Syndr 2022;14:69. 10.1186/s13098-022-00826-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Ali El Chab Parolin S, Benicio Stocco R, Kneipp Lopes JDC, et al. Association between inpatient glycemic variability and COVID-19 mortality: a prospective study. Diabetol Metab Syndr 2023;15:185. 10.1186/s13098-023-01157-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Liu P, Li Z, Tang S, et al. Association between glycemic variability and mortality in critically ill patients with heart failure. Sci Rep 2025;15:31021. 10.1038/s41598-025-16212-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Mesejo A, Montejo-González JC, Vaquerizo-Alonso C, et al. Diabetes-specific enteral nutrition formula in hyperglycemic, mechanically ventilated, critically ill patients: a prospective, open-label, blind-randomized, multicenter study. Crit Care 2015;19:390. 10.1186/s13054-015-1108-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Wang J, Wang LL, Yang YM, et al. Impact of acute glycemic variability on short-term outcomes in patients with ST-segment elevation myocardial infarction: a multicenter population-based study. Cardiovasc Diabetol 2024;23:155. 10.1186/s12933-024-02250-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sun B, Luo Z, Zhou J. Comprehensive elaboration of glycemic variability in diabetic macrovascular and microvascular complications. Cardiovasc Diabetol 2021;20:9. 10.1186/s12933-020-01200-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Lu N, Yin F. Relationship Between Hyperuricemia-Waist Phenotype and Obstructive Sleep Apnea in Type 2 Diabetes Mellitus. Diabetes Metab Syndr Obes 2023;16:1505-13. 10.2147/DMSO.S408637 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Shen H, Zhao J, Liu Y, et al. Interactions between and Shared Molecular Mechanisms of Diabetic Peripheral Neuropathy and Obstructive Sleep Apnea in Type 2 Diabetes Patients. J Diabetes Res 2018;2018:3458615. 10.1155/2018/3458615 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Wang X, Cao Y. A Narrative Review: Relationship Between Glycemic Variability and Emerging Complications of Diabetes Mellitus. Biomolecules 2025;15:188. 10.3390/biom15020188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Jaroonpipatkul S, Sathapanasiri T, Maliang C, et al. Clinical Characteristics and Prevalence of Atrial High-Rate Episodes in Patients With Cardiac Implantable Electronic Devices. Cureus 2024;16:e75380. 10.7759/cureus.75380 [DOI] [PMC free article] [PubMed] [Google Scholar]