Abstract
Objective
To examine the associations of readily available clinical indicators and the neutrophil-to-lymphocyte ratio (NLR) with moderate-to-severe obstructive sleep apnea (OSA) in patients with acute ischemic stroke and to explore the discriminatory performance of BMI, NLR, and a combination of BMI, snoring, and NLR.
Methods
This single-center retrospective study included 220 patients with acute ischemic stroke who completed overnight polysomnography during hospitalization between October 2024 and April 2026. Moderate-to-severe OSA was defined as an apnea-hypopnea index ≥15 events/h. Clinical characteristics, laboratory indicators, and polysomnographic parameters were compared between groups. Multivariable logistic regression was used to evaluate factors associated with moderate-to-severe OSA, and receiver operating characteristic analysis assessed discriminatory performance. Spearman correlation and NLR tertile analyses explored the relationships between NLR and sleep-disordered breathing and nocturnal hypoxia.
Results
Moderate-to-severe OSA was present in 109 of 220 patients (49.5%). After multivariable adjustment for prespecified clinical covariates, higher BMI (aOR 1.204, 95% CI 1.088–1.332), snoring (aOR 2.910, 95% CI 1.485–5.705), and higher NLR (aOR 2.291, 95% CI 1.449–3.622) remained associated with moderate-to-severe OSA. Modified Poisson sensitivity analyses yielded consistent associations for BMI, snoring, and NLR. The combined model had an AUC of 0.798 (95% CI, 0.736–0.851), indicating moderate discriminatory ability. Higher NLR was also associated with greater sleep-disordered breathing and nocturnal hypoxia burden.
Conclusion
BMI, snoring, and NLR were associated with moderate-to-severe OSA in this single-center acute ischemic stroke cohort. Their combined discriminatory performance was moderate. These observational findings are hypothesis-generating and require prospective multicenter validation before clinical application.
Keywords: acute ischemic stroke, obstructive sleep apnea, moderate-to-severe obstructive sleep apnea, neutrophil-to-lymphocyte ratio, body mass index, snoring
Introduction
Acute ischemic stroke remains a leading cause of death and long-term disability worldwide and accounts for the majority of stroke cases in China.1,2 Beyond acute reperfusion and secondary prevention, identifying clinically relevant comorbidities that may influence recovery and long-term vascular risk is an important component of comprehensive stroke management.
Obstructive sleep apnea (OSA) is characterized by recurrent upper-airway obstruction during sleep, resulting in intermittent hypoxia, sleep fragmentation, sympathetic activation, oxidative stress, and systemic inflammation.3,4 OSA is highly prevalent after stroke, and moderate-to-severe sleep-disordered breathing is common in this population.5 Importantly, increasing OSA severity is accompanied by greater respiratory-event and hypoxic burden, which may contribute to vascular dysfunction, impaired neurological recovery, and recurrent vascular events.4–6 Therefore, moderate-to-severe OSA represents a clinically relevant subgroup in patients with acute ischemic stroke and warrants particular attention during hospitalization.
Current stroke guidelines support evaluation or screening for sleep apnea after ischemic stroke or transient ischemic attack.7,8 However, timely objective sleep assessment may be challenging during acute hospitalization because of neurological deficits, limited mobility, impaired communication, and monitoring intolerance. Easily available clinical indicators that can help characterize patients more likely to have moderate-to-severe OSA may therefore complement, but not replace, objective sleep assessment.
Anthropometric and symptom-based variables such as BMI, neck circumference, sex, and snoring are commonly incorporated into OSA screening tools and stroke-specific prediction models, but their discriminatory performance has generally been modest and external validation remains limited.9–11In addition to these conventional clinical indicators, systemic inflammatory markers may provide complementary information. Recurrent intermittent hypoxia and sleep fragmentation in OSA can promote oxidative stress, sympathetic activation, and inflammatory responses,3,4 whereas the neutrophil-to-lymphocyte ratio (NLR) is an inexpensive and routinely available marker reflecting the balance between innate inflammatory activation and lymphocyte-related immune regulation. Previous meta-analytic evidence has shown higher NLR levels in patients with OSA than in controls, with greater differences observed in more severe OSA, supporting the potential relevance of NLR as a marker of OSA-related systemic inflammation.12 However, the extent to which NLR provides clinically relevant information alongside readily available anthropometric and symptom-based indicators in patients with acute ischemic stroke remains insufficiently characterized.
Accordingly, this study aimed to examine readily available anthropometric, symptom-based, and inflammatory indicators associated with moderate-to-severe OSA in patients with acute ischemic stroke and to explore the discriminatory performance of selected associated factors. We hypothesized that BMI, snoring, and NLR would remain associated with moderate-to-severe OSA after adjustment for clinically relevant covariates, including stroke severity. Given the retrospective observational design, these analyses were intended to characterize associations and generate hypotheses for future prospective validation rather than to establish causal relationships or a definitive clinical prediction tool.
Methods
Study Population
This was a single-center retrospective study designed and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement.13 The completed STROBE checklist is provided in Supplementary Material 1. Patients with acute ischemic stroke who were hospitalized in the Department of Neurology, Affiliated Hospital of Yangzhou University, and completed overnight polysomnography (PSG) between October 2024 and April 2026 were included. Acute ischemic stroke was diagnosed according to the 2023 updated guidelines for the clinical management of ischemic cerebrovascular diseases issued by the Chinese Stroke Association and relevant neuroimaging findings. Specifically, eligible patients had acute focal neurological deficits, with cranial computed tomography or magnetic resonance imaging findings supporting the diagnosis of acute ischemic stroke.14 This study was approved by the Ethics Committee of Yangzhou University, with the approval number 2023-YKL09. Because this was a retrospective study and all data were anonymized, written informed consent had been obtained from all patients or their legally authorized representatives.
The inclusion criteria were as follows: (1) age ≥18 years; (2) diagnosis of acute ischemic stroke confirmed by cranial computed tomography or magnetic resonance imaging; (3) completion of sleep monitoring during hospitalization, with available sleep monitoring data for calculating the apnea-hypopnea index (AHI); and (4) complete data on baseline characteristics, clinical symptoms, laboratory indicators, and sleep monitoring parameters. The exclusion criteria were as follows: (1) transient ischemic attack, intracerebral hemorrhage, subarachnoid hemorrhage, or other non-ischemic cerebrovascular diseases; (2) severe disturbance of consciousness, clinically evident acute infection at admission, severe cardiopulmonary insufficiency, end-stage malignancy, or other conditions that could affect the completion of sleep monitoring or interpretation of monitoring results; (3) previously diagnosed OSA treated with continuous positive airway pressure, oral appliances, or related surgical interventions; and (4) insufficient sleep monitoring duration, poor monitoring data quality, or missing key variables. Clinically evident acute infection at admission was assessed based on a combined review of the admission diagnosis, body temperature, infection-related symptoms and signs, complete blood count, C-reactive protein (CRP), and chest imaging. Patients considered to have acute infection on this combined clinical assessment were excluded.
During the study period, 352 patients were initially assessed for eligibility. Before overnight polysomnography (PSG), 96 patients were excluded because of non-ischemic cerebrovascular disease or transient ischemic attack (n=22), severe disturbance of consciousness, severe cardiopulmonary insufficiency, end-stage malignancy, or other conditions precluding PSG (n=31), previously treated OSA (n=7), or refusal or inability to tolerate or complete PSG (n=36). A total of 256 patients underwent overnight PSG during hospitalization. After PSG and data review, a further 36 patients were excluded because of insufficient monitoring duration (n=12), poor monitoring data quality (n=9), or missing key baseline, laboratory, or sleep-monitoring data (n=15). Finally, 220 patients were included in the analysis, including 111 patients with an AHI <15 events/h and 109 patients with an AHI ≥15 events/h. The detailed participant selection process is shown in Figure 1.
Figure 1.
Flowchart of patient inclusion.
Abbreviation: OSA, obstructive sleep apnea.
Diagnosis and Grouping Criteria for OSA
All patients underwent overnight polysomnography (PSG) within 7 days after stroke onset during hospitalization using the SOMNOmedics V6 polysomnography system (SOMNOmedics GmbH, Germany). The median interval from stroke onset to PSG was 4 days (range, 1–7 days), with most examinations performed on days 4–5 after stroke onset. Thus, all PSG examinations were conducted during the acute post-stroke phase. The monitoring parameters included electroencephalography, electrooculography, chin electromyography, oronasal airflow, thoracoabdominal respiratory movement, oxygen saturation, heart rate, body position, and sleep staging. The use of overnight PSG for the diagnostic evaluation of OSA was consistent with the American Academy of Sleep Medicine (AASM) clinical practice guideline for diagnostic testing for adult OSA.15 Sleep stages and sleep-related respiratory events were scored according to the AASM Manual for the Scoring of Sleep and Associated Events, Version 3 (2023).16 Automated scoring generated by the SOMNOmedics system was subsequently manually reviewed and corrected by sleep technologists, with final review by an experienced physician in the sleep center. PSG scoring was completed as part of routine clinical assessment before study-group classification and statistical analysis; therefore, the scorers were unaware of the subsequent study-group assignment.
AHI was defined as the number of apnea and hypopnea events per hour of sleep and was used as the primary criterion for grading OSA severity. OSA severity was classified as no OSA (AHI <5 events/h), mild OSA (5≤AHI<15 events/h), moderate OSA (15≤AHI<30 events/h), and severe OSA (AHI ≥30 events/h).16 The present study specifically focused on moderate-to-severe OSA because the primary objective was to examine clinical and inflammatory indicators associated with this clinically established severity category. Accordingly, an AHI ≥15 events/h was used to define moderate-to-severe OSA and coded as 1, whereas an AHI <15 events/h was coded as 0. We acknowledge that dichotomization does not preserve the full ordinal or continuous information contained in AHI; however, this approach was retained because it directly corresponded to the prespecified outcome of interest rather than being used to characterize the full dose–response relationship across OSA severity.
Data Collection
Patient data were retrospectively collected from the hospital electronic medical record system, laboratory information system, and sleep monitoring report system. The collected data included baseline characteristics, medical history, sleep-related respiratory symptoms, STOP-Bang scores, laboratory indicators, and sleep monitoring parameters. All data were extracted and checked by the researchers using a standardized data collection form.
Baseline and disease-related variables included sex, age, height, body weight, body mass index (BMI), neck circumference, smoking history, alcohol consumption, hypertension, diabetes mellitus, coronary heart disease, atrial fibrillation, hyperlipidemia, admission National Institutes of Health Stroke Scale (NIHSS) score, and admission modified Rankin Scale (mRS) score. BMI was calculated as body weight in kilograms divided by height in meters squared. Previous medical histories, including hypertension, diabetes mellitus, coronary heart disease, atrial fibrillation, and hyperlipidemia, were determined according to previous diagnostic records, medication history, or admission diagnoses. Stroke severity at admission was assessed using the National Institutes of Health Stroke Scale (NIHSS), with higher scores indicating greater neurological deficit severity. Functional status at admission was assessed using the modified Rankin Scale (mRS), with higher scores indicating greater functional disability.
Sleep-related respiratory symptoms included snoring, daytime fatigue or sleepiness, and witnessed apnea. For patients with limited verbal expression, impaired consciousness, or an inability to provide an accurate medical history, symptom information was determined based on descriptions from family members or caregivers, together with medical record documentation.
The STOP-Bang questionnaire was used to assess the risk of OSA. This questionnaire consists of eight items: snoring, tiredness, observed apnea, high blood pressure, BMI, age, neck circumference, and sex. Each item is scored as 0 or 1, with a total score ranging from 0 to 8; higher scores indicate a higher risk of OSA.17 In this study, the total STOP-Bang score was collected and used for between-group comparisons and univariate logistic regression analysis. Because the total STOP-Bang score overlaps with its component items, particularly age, sex, BMI, neck circumference, snoring, and hypertension, it was not entered into the final multivariable logistic regression model together with these component variables to avoid duplicate adjustment and potential collinearity.
Laboratory indicators were obtained from fasting venous blood samples collected in the morning on the day after admission and before overnight PSG. The first complete blood count and relevant laboratory test results after admission were used for the present analysis, with particular attention to absolute neutrophil and lymphocyte counts. The neutrophil-to-lymphocyte ratio (NLR) was calculated as the absolute neutrophil count divided by the absolute lymphocyte count. Missing, abnormal, or inconsistent laboratory data were verified using the original laboratory reports and medical records.
Sleep monitoring parameters included total sleep time, sleep efficiency, sleep onset latency, REM latency, proportions of N1, N2, N3, and REM sleep, supine time, AHI, numbers of obstructive apnea, central apnea, mixed apnea, and hypopnea events, REM AHI, NREM AHI, supine AHI, mean oxygen saturation during sleep, minimum oxygen saturation during NREM sleep, minimum oxygen saturation during REM sleep, oxygen desaturation index ≥3%, time with oxygen saturation below 90%, and number of awakenings during sleep. These parameters were mainly used to describe sleep-disordered breathing burden, nocturnal hypoxia burden, and sleep fragmentation in the two groups. Because sleep monitoring parameters were highly related to the diagnosis and severity classification of moderate-to-severe OSA, they were not included in the main multivariable logistic regression model to avoid overlap between the outcome definition and explanatory variables.
Statistical Analysis
Statistical analyses were performed using IBM SPSS Statistics version 32.0. Because this was a retrospective study including all eligible patients during the predefined study period, no prospective sample size calculation was performed. The final cohort comprised 220 patients, including 109 patients with moderate-to-severe OSA. Model complexity was restricted to a clinically prespecified and parsimonious set of covariates relative to the number of outcome events to reduce the risk of overfitting. The distributions of continuous variables were formally assessed using the Shapiro–Wilk test. Normally distributed continuous variables were expressed as mean ± standard deviation and compared between groups using the independent-samples t test. Non-normally distributed continuous variables were expressed as median and interquartile range and compared using the Mann–Whitney U-test. Categorical variables were expressed as frequencies and percentages and compared using the chi-square test or Fisher’s exact test, as appropriate.
The presence of moderate-to-severe OSA was used as the dependent variable, with non-moderate-to-severe OSA coded as 0 and moderate-to-severe OSA coded as 1. Univariate logistic regression analyses were performed to describe the crude associations between candidate baseline variables and moderate-to-severe OSA. Candidate variables included age, sex, BMI, neck circumference >40 cm, snoring, hypertension, diabetes mellitus, atrial fibrillation, coronary heart disease, hyperlipidemia, alcohol consumption, smoking, witnessed apnea, daytime fatigue or sleepiness, PSQI score, MoCA score, STOP-Bang score, C-reactive protein, neutrophil count, lymphocyte count, NLR, admission NIHSS score, and admission mRS score. Univariate statistical significance was not used as the sole criterion for multivariable model construction. BMI, snoring, and NLR were considered the primary variables of interest, whereas age, sex, neck circumference >40 cm, hypertension, diabetes mellitus, atrial fibrillation, smoking, alcohol consumption, hyperlipidemia, and admission NIHSS score were included as clinically relevant adjustment covariates based on previous evidence, clinical plausibility, and model parsimony. The total STOP-Bang score was not included in the same multivariable model because of substantial overlap with several of its component variables. Multivariable logistic regression was performed using the Enter method, and admission mRS score was additionally included in a sensitivity analysis. Because moderate-to-severe OSA was common in this cohort, modified Poisson regression with robust variance was also performed as a sensitivity analysis using the same primary covariate set to directly estimate adjusted prevalence ratios (aPRs).18 Logistic regression results are presented as adjusted odds ratios (aORs), and modified Poisson regression results as aPRs, with corresponding 95% confidence intervals (CIs). All tests were two-sided, and P<0.05 was considered statistically significant.
Receiver operating characteristic (ROC) curve analysis was performed to evaluate the discriminatory ability of BMI, NLR, and the combined model for identifying moderate-to-severe OSA. The combined model was constructed based on multivariable logistic regression using BMI, snoring, and NLR. The area under the ROC curve (AUC), 95% confidence interval (CI), optimal cutoff value based on the Youden index, sensitivity, and specificity were calculated.
To further explore the relationships between NLR and indicators of sleep-disordered breathing burden and nocturnal hypoxia, Spearman rank correlation analysis was used to assess the correlations of NLR with REM AHI, NREM AHI, supine AHI, oxygen desaturation index ≥3%, time with oxygen saturation <90%, mean oxygen saturation during sleep, minimum oxygen saturation during NREM sleep, and minimum oxygen saturation during REM sleep. Patients were divided into low, middle, and high tertile groups according to NLR. The chi-square test was used to compare the proportion of moderate-to-severe OSA across NLR tertiles, and the linear-by-linear association test was used to evaluate the trend in the proportion of moderate-to-severe OSA with increasing NLR levels.
Results
Comparison of Baseline Characteristics, Clinical Symptoms, and Scale Scores
A total of 220 patients with acute ischemic stroke were included in this study, including 111 patients in the non-moderate-to-severe OSA group and 109 patients in the moderate-to-severe OSA group. The patient inclusion flowchart is shown in Figure 1. There were no significant between-group differences in admission NIHSS score (median [IQR], 4.0 [2.5–6.0] vs 4.0 [3.0–6.0]; P=0.307) or admission mRS score (2.0 [1.0–3.0] vs 2.0 [1.0–3.0]; P=0.945). As shown in Table 1, there were also no significant differences between the two groups in age, hypertension, diabetes mellitus, atrial fibrillation, coronary heart disease, hyperlipidemia, alcohol consumption, smoking, witnessed apnea, daytime fatigue or sleepiness, PSQI score, MoCA score, or C-reactive protein level (all P>0.05). Compared with the non-moderate-to-severe OSA group, the moderate-to-severe OSA group had significantly higher proportions of male sex, neck circumference >40 cm, and snoring, as well as higher BMI, STOP-Bang score, neutrophil count, and NLR, but a lower lymphocyte count (all P<0.05). Detailed results are presented in Table 1.
Table 1.
Comparison of Baseline Characteristics and Scale Scores Between Groups
| Variable | Non-Moderate-to-Severe OSA Group (n=111) | Moderate-to-Severe OSA Group (n=109) | P value |
|---|---|---|---|
| Age, years | 59.51 ± 13.07 | 59.49 ± 12.18 | 0.987 |
| Admission NIHSS score | 4.0 (2.5–6.0) | 4.0 (3.0–6.0) | 0.307 |
| Admission mRS score | 2.0 (1.0–3.0) | 2.0 (1.0–3.0) | 0.945 |
| Male sex, n (%) | 56 (50.5) | 76 (69.7) | 0.004 |
| BMI, kg/m2 | 23.20 (21.30–25.65) | 26.20 (24.20–29.40) | <0.001 |
| Hypertension, n (%) | 47 (42.3) | 55 (50.5) | 0.227 |
| Diabetes mellitus, n (%) | 24 (21.6) | 19 (17.4) | 0.433 |
| Atrial fibrillation, n (%) | 7 (6.3) | 6 (5.5) | 0.801 |
| Coronary heart disease, n (%) | 5 (4.5) | 10 (9.2) | 0.169 |
| Hyperlipidemia, n (%) | 42 (37.8) | 50 (45.9) | 0.227 |
| Alcohol consumption, n (%) | 28 (25.2) | 30 (27.5) | 0.699 |
| Smoking, n (%) | 32 (28.8) | 35 (32.1) | 0.597 |
| Neck circumference >40 cm, n (%) | 29 (26.1) | 56 (51.4) | <0.001 |
| Snoring, n (%) | 36 (32.4) | 66 (60.6) | <0.001 |
| Witnessed apnea, n (%) | 24 (21.6) | 25 (22.9) | 0.815 |
| Daytime fatigue/sleepiness, n (%) | 40 (36.0) | 38 (34.9) | 0.856 |
| PSQI score | 8.0 (5.5–11.5) | 8.0 (5.0–11.0) | 0.575 |
| MoCA score | 20.0 (15.5–24.5) | 20.0 (16.0–25.0) | 0.990 |
| STOP-Bang score | 3.0 (2.0–4.0) | 4.0 (3.0–5.0) | <0.001 |
| C-reactive protein, mg/L | 1.24 (0.50–3.23) | 1.24 (0.50–3.25) | 0.647 |
| Neutrophil count, ×109/L | 3.78 (2.97–5.26) | 4.47 (3.78–6.05) | <0.001 |
| Lymphocyte count, ×109/L | 5.05 (3.42–6.06) | 4.22 (2.61–5.44) | 0.003 |
| NLR | 0.84 (0.59–1.22) | 1.21 (0.85–1.93) | <0.001 |
Notes: Continuous variables were assessed for normality using the Shapiro–Wilk test. Normally distributed continuous variables are presented as mean ± standard deviation and were compared using the independent-samples t test. Non-normally distributed continuous variables are presented as median (interquartile range [IQR]) and were compared using the Mann–Whitney U-test. Categorical variables are presented as n (%) and were compared using the chi-square test or Fisher’s exact test, as appropriate. P<0.05 was considered statistically significant.
Abbreviations: OSA, obstructive sleep apnea; NIHSS, National Institutes of Health Stroke Scale; mRS, modified Rankin Scale; BMI, body mass index; PSQI, Pittsburgh Sleep Quality Index; MoCA, Montreal Cognitive Assessment; CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio.
Comparison of Sleep Monitoring Parameters Between Groups
As shown in Table 2, there were no significant differences between the two groups in total sleep time, sleep efficiency, sleep onset latency, REM latency, proportions of N1, N2, and REM sleep, supine time, or number of awakenings during sleep (all P>0.05).
Table 2.
Comparison of Sleep Monitoring Parameters Between Groups
| Variable | Non-Moderate-to-Severe OSA Group (n=111) | Moderate-to-Severe OSA Group (n=109) | P value |
|---|---|---|---|
| Total sleep time, min | 447.0 (374.2–488.2) | 416.4 (352.3–473.7) | 0.104 |
| Sleep efficiency, % | 84.7 (71.8–92.6) | 83.3 (70.4–89.9) | 0.557 |
| Sleep onset latency, min | 5.2 (1.9–12.8) | 3.9 (1.1–10.4) | 0.148 |
| REM latency, min | 101.5 (68.0–178.7) | 99.5 (60.5–164.5) | 0.729 |
| N1 stage, % | 3.8 (2.1–7.7) | 5.9 (2.6–9.2) | 0.072 |
| N2 stage, % | 56.73 ± 13.39 | 59.89 ± 14.07 | 0.089 |
| N3 stage, % | 19.5 (9.5–30.6) | 16.1 (8.0–24.9) | 0.024 |
| REM stage, % | 17.3 (11.4–23.0) | 15.3 (8.2–22.4) | 0.179 |
| Obstructive apnea events, n | 5.0 (1.0–17.0) | 32.0 (7.0–96.0) | <0.001 |
| Mixed apnea events, n | 0.0 (0.0–0.0) | 0.0 (0.0–4.0) | <0.001 |
| Central apnea events, n | 0.0 (0.0–1.0) | 0.0 (0.0–1.0) | 0.786 |
| Hypopnea events, n | 50.0 (21.0–82.0) | 168.0 (116.0–241.0) | <0.001 |
| REM AHI, events/h | 11.1 (2.8–19.0) | 35.1 (25.3–52.3) | <0.001 |
| NREM AHI, events/h | 5.4 (1.9–10.1) | 34.9 (26.0–51.4) | <0.001 |
| Supine time, min | 166.7 (0.0–290.1) | 158.4 (0.0–286.5) | 0.781 |
| Supine AHI, events/h | 7.2 (0.3–14.5) | 30.0 (0.0–51.6) | <0.001 |
| Mean oxygen saturation during sleep, % | 95.0 (94.0–96.0) | 94.0 (93.0–95.0) | <0.001 |
| Minimum oxygen saturation during NREM sleep, % | 88.0 (84.0–91.0) | 82.0 (75.0–87.0) | <0.001 |
| Minimum oxygen saturation during REM sleep, % | 90.0 (86.0–92.0) | 85.0 (77.0–89.0) | <0.001 |
| Oxygen desaturation index ≥3%, events/h | 8.5 (3.8–14.1) | 38.7 (23.5–55.2) | <0.001 |
| Time with oxygen saturation <90%, h | 0.00 (0.00–0.10) | 0.30 (0.10–1.20) | <0.001 |
| Awakenings during sleep, n | 242.0 (139.0–453.0) | 304.0 (193.0–494.0) | 0.067 |
Notes: Continuous variables were assessed for normality using the Shapiro–Wilk test. Normally distributed continuous variables are presented as mean ± standard deviation and were compared using the independent-samples t test. Non-normally distributed continuous variables are presented as median (interquartile range [IQR]) and were compared using the Mann–Whitney U-test. P<0.05 was considered statistically significant.
Abbreviations: OSA, obstructive sleep apnea; REM, rapid eye movement; NREM, non-rapid eye movement; AHI, apnea-hypopnea index.
Compared with the non-moderate-to-severe OSA group, the moderate-to-severe OSA group had a significantly lower proportion of N3 sleep and significantly higher numbers of obstructive apnea events, mixed apnea events, and hypopnea events. REM AHI, NREM AHI, and supine AHI were also significantly higher in the moderate-to-severe OSA group. In addition, this group had lower mean oxygen saturation during sleep, lower minimum oxygen saturation during NREM and REM sleep, a higher oxygen desaturation index ≥3%, and a longer duration of oxygen saturation <90% (all P<0.05). Detailed results are shown in Table 2.
Univariate Logistic Regression Analysis of Factors Associated with Moderate-to-Severe OSA
Univariate logistic regression analyses were performed for the candidate baseline variables specified in the Methods. Male sex, higher BMI, neck circumference >40 cm, snoring, higher STOP-Bang score, higher neutrophil count, lower lymphocyte count, and higher NLR were significantly associated with moderate-to-severe OSA in the unadjusted analyses (all P<0.05). Detailed results for all candidate variables are provided in Supplementary Table S1.
Multivariable Regression Analyses of Factors Associated with Moderate-to-Severe OSA
In the primary multivariable logistic regression model, higher BMI (aOR 1.204, 95% CI 1.088–1.332; P<0.001), snoring (aOR 2.910, 95% CI 1.485–5.705; P=0.002), and higher NLR (aOR 2.291, 95% CI 1.449–3.622; P<0.001) remained associated with moderate-to-severe OSA after adjustment for the prespecified covariates (Table 3 and Figure 2A). Additional adjustment for admission mRS score did not materially alter these associations. In the modified Poisson regression sensitivity analysis using the same primary covariate set, BMI (aPR 1.063, 95% CI 1.027–1.100; P<0.001), snoring (aPR 1.519, 95% CI 1.154–1.998; P=0.003), and NLR (aPR 1.118, 95% CI 1.026–1.217; P=0.010) remained significantly associated with moderate-to-severe OSA (Figure 2B and Supplementary Table S4). Male sex (aPR 1.449, 95% CI 1.059–1.984; P=0.021) and hyperlipidemia (aPR 1.309, 95% CI 1.020–1.678; P=0.034) also reached statistical significance in the modified Poisson model, whereas the remaining covariates did not.
Table 3.
Multivariable Logistic Regression Analysis of Factors Associated with Moderate-to-Severe OSA
| Variable | B | SE | Wald χ2 | Adjusted OR | 95% CI | P value |
|---|---|---|---|---|---|---|
| Age | −0.002 | 0.014 | 0.027 | 0.998 | 0.970–1.026 | 0.869 |
| Male sex | 0.594 | 0.374 | 2.527 | 1.812 | 0.871–3.770 | 0.112 |
| BMI | 0.186 | 0.052 | 12.920 | 1.204 | 1.088–1.332 | <0.001 |
| Neck circumference >40 cm | 0.425 | 0.374 | 1.293 | 1.530 | 0.735–3.182 | 0.255 |
| Snoring | 1.068 | 0.343 | 9.676 | 2.910 | 1.485–5.705 | 0.002 |
| Hypertension | 0.133 | 0.350 | 0.145 | 1.143 | 0.575–2.270 | 0.704 |
| Diabetes mellitus | −0.550 | 0.450 | 1.496 | 0.577 | 0.239–1.393 | 0.221 |
| Atrial fibrillation | 0.716 | 0.719 | 0.989 | 2.045 | 0.499–8.378 | 0.320 |
| Smoking | −0.036 | 0.438 | 0.007 | 0.964 | 0.409–2.274 | 0.934 |
| Alcohol consumption | −0.255 | 0.427 | 0.357 | 0.775 | 0.336–1.789 | 0.550 |
| Hyperlipidemia | 0.490 | 0.345 | 2.018 | 1.633 | 0.830–3.211 | 0.155 |
| NLR | 0.829 | 0.234 | 12.591 | 2.291 | 1.449–3.622 | <0.001 |
| Admission NIHSS score | 0.055 | 0.072 | 0.586 | 1.057 | 0.917–1.217 | 0.444 |
Notes: The dependent variable was moderate-to-severe obstructive sleep apnea, coded as 0 = non-moderate-to-severe OSA and 1 = moderate-to-severe OSA. The primary multivariable logistic regression model included age, sex, BMI, neck circumference >40 cm, snoring, hypertension, diabetes mellitus, atrial fibrillation, smoking, alcohol consumption, hyperlipidemia, NLR, and admission NIHSS score. For categorical variables, female sex and absence/no were used as the reference categories. Adjusted OR indicates the association between each factor and moderate-to-severe OSA after adjustment for the other covariates in the model.
Abbreviations: OSA, obstructive sleep apnea; BMI, body mass index; NLR, neutrophil-to-lymphocyte ratio; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; CI, confidence interval; SE, standard error.
Figure 2.
Multivariable-adjusted associations with moderate-to-severe obstructive sleep apnea estimated using logistic and modified Poisson regression. (A) presents adjusted odds ratios (aORs) and 95% confidence intervals (CIs) from the primary multivariable logistic regression model. (B) presents adjusted prevalence ratios (aPRs) and 95% CIs from the modified Poisson regression model with robust variance used as a sensitivity analysis. Both models included age, sex, BMI, neck circumference >40 cm, snoring, hypertension, diabetes mellitus, atrial fibrillation, smoking, alcohol consumption, hyperlipidemia, NLR, and admission NIHSS score. Squares indicate point estimates and horizontal lines indicate 95% CIs; the vertical dashed line represents an effect estimate of 1.0.
Abbreviations: OSA, obstructive sleep apnea; BMI, body mass index; NLR, neutrophil-to-lymphocyte ratio; NIHSS, National Institutes of Health Stroke Scale; aOR, adjusted odds ratio; aPR, adjusted prevalence ratio; CI, confidence interval.
Association of NLR with Sleep-Disordered Breathing Burden and Nocturnal Hypoxia-Related Indicators
To further examine the relationship between NLR and indicators of sleep-disordered breathing burden and nocturnal hypoxia, supplementary Spearman correlation analyses were performed. NLR was positively correlated with REM AHI, NREM AHI, supine AHI, oxygen desaturation index ≥3%, and time with oxygen saturation <90%. It was negatively correlated with mean oxygen saturation during sleep, minimum oxygen saturation during NREM sleep, and minimum oxygen saturation during REM sleep. All correlations were statistically significant (all P<0.05). Detailed correlation coefficients are provided in Supplementary Table S2.
After patients were divided into low, middle, and high tertile groups according to NLR, the proportions of moderate-to-severe OSA were 27.4%, 56.8%, and 64.4%, respectively. The chi-square test showed a significant difference in the proportion of moderate-to-severe OSA across NLR tertiles (χ2=22.294, P<0.001). The linear-by-linear association test further showed an increasing trend in the proportion of moderate-to-severe OSA with higher NLR tertiles (χ2=19.883, P for trend<0.001). Detailed tertile data are provided in Supplementary Table S3 and summarized visually in Figure 3.
Figure 3.
Proportion of moderate-to-severe OSA across NLR tertiles. Data are presented as proportions with 95% Wilson confidence intervals. P for trend was calculated using the linear-by-linear association test.
Abbreviations: NLR, neutrophil-to-lymphocyte ratio; OSA, obstructive sleep apnea.
Discriminatory Performance of BMI, NLR, and the Combined Model
Receiver operating characteristic (ROC) curve analysis demonstrated that BMI and NLR had modest discriminatory value for identifying moderate-to-severe OSA, with AUCs of 0.736 (95% CI, 0.669–0.796) and 0.684 (95% CI, 0.610–0.753), respectively. A combined model incorporating BMI, snoring, and NLR achieved a higher AUC of 0.798 (95% CI, 0.736–0.851). At the optimal cutoff value of 0.340, the combined model yielded a sensitivity of 88.1% and a specificity of 55.9%. Detailed ROC results are presented in Table 4 and Figure 4.
Table 4.
ROC Analysis of BMI, NLR, and the Combined Model for Identifying Moderate-to-Severe OSA
| Indicator | AUC | 95% CI | Optimal Cutoff | Sensitivity (%) | Specificity (%) |
|---|---|---|---|---|---|
| BMI | 0.736 | 0.669–0.796 | 23.1 | 87.2 | 47.7 |
| NLR | 0.684 | 0.610–0.753 | 0.751 | 84.4 | 46.8 |
| Combined model (BMI + snoring + NLR) | 0.798 | 0.736–0.851 | 0.340 | 88.1 | 55.9 |
Notes: The combined model was constructed using BMI, snoring, and NLR.
Abbreviations: ROC, receiver operating characteristic; AUC, area under the curve; CI, confidence interval; BMI, body mass index; NLR, neutrophil-to-lymphocyte ratio; OSA, obstructive sleep apnea.
Figure 4.
Receiver operating characteristic curves of BMI, NLR, and the combined model for identifying moderate-to-severe OSA. Receiver operating characteristic curves were used to evaluate the discriminatory ability of BMI, NLR, and the combined model incorporating BMI, snoring, and NLR for identifying moderate-to-severe obstructive sleep apnea in patients with acute ischemic stroke.
Abbreviations: ROC, receiver operating characteristic; BMI, body mass index; NLR, neutrophil-to-lymphocyte ratio; OSA, obstructive sleep apnea; AUC, area under the curve; CI, confidence interval.
Discussion
In this single-center retrospective study of 220 patients with acute ischemic stroke, higher BMI, snoring, and NLR remained associated with moderate-to-severe OSA after adjustment for prespecified clinical covariates in the primary logistic regression model. These three associations were also retained in the modified Poisson sensitivity analysis, although the effect estimates were attenuated when expressed as prevalence ratios rather than odds ratios. Male sex and hyperlipidemia additionally reached statistical significance in the modified Poisson model but not in the primary logistic model, suggesting that these secondary covariate-specific findings were model-dependent and should be interpreted cautiously. Admission NIHSS and mRS scores did not differ significantly between groups, and additional adjustment for admission mRS did not materially alter the main associations. Rather than representing a new diagnostic or prediction model, these findings mainly confirm and extend existing evidence on anthropometric, symptom-based, and inflammatory correlates of OSA within an acute ischemic stroke population.9,12,19
Higher BMI remained associated with moderate-to-severe OSA after multivariable adjustment, consistent with the well-established relationship between obesity and OSA.20–22 Obesity may contribute to upper-airway narrowing and collapsibility through both overall and regional fat accumulation.23–25 In the present cohort, BMI remained associated with moderate-to-severe OSA whereas neck circumference >40 cm did not retain statistical significance after adjustment. This difference should not be interpreted as evidence that neck circumference is unimportant; rather, BMI and neck circumference capture overlapping aspects of obesity-related risk, and the use of a single neck-circumference threshold may provide less information than a continuous measure of overall adiposity. Thus, our findings primarily confirm the relevance of obesity-related characteristics in patients with acute ischemic stroke rather than establish BMI as a stand-alone screening or diagnostic indicator.
Snoring remained associated with moderate-to-severe OSA after multivariable adjustment, consistent with previous evidence identifying snoring as a relevant component of OSA screening in both general and stroke populations.19,21,22 In patients with stroke or transient ischemic attack, modified STOP-Bang-based approaches combined with nocturnal oxygen desaturation have shown useful but imperfect discrimination for OSA,19 while a recent systematic review of stroke-specific OSA prediction models highlighted substantial methodological heterogeneity and limited external validation.9 Therefore, the present findings should not be interpreted as establishing a new prediction model; rather, they support the continued relevance of a simple symptom such as snoring within an acute ischemic stroke population. Because symptom reporting may be limited after stroke, information from family members or caregivers may provide useful complementary information, but snoring should be regarded as a clinical clue prompting objective sleep assessment rather than as a diagnostic indicator.
Higher NLR remained associated with moderate-to-severe OSA after multivariable adjustment. This finding is consistent with previous evidence showing higher NLR levels in patients with OSA.12 Recurrent intermittent hypoxia and sleep fragmentation may contribute to oxidative stress, sympathetic activation, endothelial dysfunction, and systemic inflammatory responses.26 However, interpretation of NLR is particularly complex in patients with acute ischemic stroke because stroke itself induces substantial inflammatory and immune responses, and NLR has also been associated with stroke severity and prognosis.27–29 In the present study, patients with clinically evident acute infection were excluded and blood sampling conditions were standardized; nevertheless, NLR may still reflect a combination of OSA-related hypoxic burden, acute stroke-related inflammation, physiological stress, medication use, and underlying comorbidities. The observed correlations between NLR and indices of sleep-disordered breathing and nocturnal hypoxia, together with the increasing proportion of moderate-to-severe OSA across NLR tertiles, provide supportive but exploratory evidence of this association. Because of the retrospective design and the use of a single NLR measurement, these findings should not be interpreted as evidence that nocturnal hypoxia causes an increase in NLR or that NLR contributes causally to the development of OSA. Prospective studies incorporating serial inflammatory measurements are needed to clarify the temporal relationship between inflammatory changes and sleep-disordered breathing after acute ischemic stroke.
The exploratory ROC analysis showed that the combination of BMI, snoring, and NLR had moderate discriminatory ability for moderate-to-severe OSA, with an AUC of 0.798 (95% CI, 0.736–0.851). At the selected cutoff, sensitivity was relatively high (88.1%), whereas specificity was modest (55.9%), indicating that the three-variable combination should not be regarded as a diagnostic tool. Previous stroke-specific OSA screening and prediction approaches have also shown variable discriminatory performance and limited external validation.9,19 Therefore, the present ROC findings should be considered exploratory and hypothesis-generating. They suggest that combining readily available anthropometric, symptom-based, and inflammatory information may warrant further evaluation, but prospective model development, calibration, internal validation, and external validation would be required before any clinical implementation.
Male sex and neck circumference >40 cm showed associations with moderate-to-severe OSA in unadjusted analyses but were not statistically significant in the fully adjusted multivariable model. This should not be interpreted as evidence that these established OSA-related characteristics are unimportant; rather, their associations may partly overlap with BMI, snoring, and other correlated clinical factors.20–22 The STOP-Bang score was also higher in patients with moderate-to-severe OSA and was associated with the outcome in univariate analysis, consistent with its established role as an OSA screening instrument.30,31 However, because the total STOP-Bang score incorporates age, sex, BMI, neck circumference, snoring, and hypertension, it was not entered together with these component variables in the multivariable model to avoid redundant adjustment and potential collinearity. Thus, the STOP-Bang findings should be interpreted as supportive descriptive evidence rather than as an additional independent factor in the present analysis.
As expected from the AHI-based group definition, patients with moderate-to-severe OSA showed a greater respiratory-event burden and more pronounced nocturnal hypoxia than those without moderate-to-severe OSA. These differences are consistent with the established physiological features of more severe OSA, including recurrent respiratory events, intermittent hypoxia, and sleep fragmentation.26,32,33 The higher oxygen desaturation index and greater nocturnal hypoxic burden observed in this group similarly reflect greater OSA severity.34 A lower proportion of N3 sleep was also observed, although this secondary finding should be interpreted cautiously given the multiple sleep parameters examined and the potential influence of the acute inpatient environment. Importantly, PSG-derived respiratory and hypoxia variables were not entered into the main regression model because they are intrinsically related to the AHI-based definition and severity of the outcome, and their inclusion could result in circular interpretation.
From a clinical perspective, the present findings should be interpreted as hypothesis-generating rather than practice-changing. BMI, snoring, and NLR are readily available during hospitalization, but their combined discriminatory performance was only moderate and they should not replace objective sleep assessment. Previous work has shown that structured diagnostic pathways can increase OSA testing after stroke or transient ischemic attack;35 however, the present study did not evaluate whether incorporating BMI, snoring, or NLR into such pathways improves diagnostic yield, patient outcomes, or resource allocation. Prospective studies are therefore needed to determine whether these readily available clinical indicators provide incremental value beyond established screening approaches before they are incorporated into clinical decision-making.
Several limitations should be acknowledged. First, this was a single-center retrospective study of hospitalized patients, and only patients who were able to complete overnight PSG and had sufficiently complete data were included. Patients who could not tolerate or complete PSG were excluded, which may have introduced selection bias and limits the generalizability of the findings to the broader population of patients with acute ischemic stroke. Second, the retrospective collection of clinical information may be subject to information bias. In particular, symptoms such as snoring, witnessed apnea, and daytime fatigue or sleepiness were obtained from medical records and, when necessary, from patients or caregivers, and may therefore be affected by incomplete documentation or recall. Third, although admission NIHSS was included in the primary multivariable model and admission mRS was additionally considered in sensitivity analysis, infarct size and infarct location were not consistently available using standardized definitions across all patients and therefore could not be included in the analyses. Residual confounding from these and other unmeasured or incompletely characterized stroke-related factors cannot be excluded. Fourth, although patients with clinically evident acute infection were excluded and blood sampling conditions were standardized, NLR was measured at a single time point during the acute post-stroke phase and may still have been influenced by acute stroke-related inflammation, physiological stress, medication use, and underlying comorbidities. Fifth, multiple secondary PSG parameters and exploratory inflammatory analyses were examined without formal adjustment for multiple comparisons; therefore, some secondary findings may represent chance associations and should be interpreted cautiously. Sixth, despite restricting the multivariable analysis to clinically relevant covariates, the sample size and number of outcome events were limited relative to the number of variables considered, and some degree of model overfitting cannot be excluded. Furthermore, the BMI–snoring–NLR combination was evaluated only through exploratory ROC analysis; formal prediction-model development, calibration assessment, internal validation, and external validation were not performed. Consequently, its discriminatory performance should not be interpreted as evidence of clinical predictive utility. Finally, PSG was performed during acute hospitalization, and sleep architecture and respiratory parameters may have been influenced by the inpatient environment, acute physical condition, medication use, body position, and nighttime clinical care. Future multicenter prospective studies with larger samples, repeated inflammatory measurements, standardized sleep assessment, and appropriate model validation are needed to confirm these findings.
Conclusion
In this single-center retrospective study of patients with acute ischemic stroke, higher BMI, snoring, and higher NLR were associated with moderate-to-severe OSA after adjustment for prespecified clinical covariates. The combination of BMI, snoring, and NLR showed moderate discriminatory ability; however, this study was not designed to develop or validate a clinical prediction model. These observational findings should therefore be regarded as hypothesis-generating and require prospective multicenter validation before any clinical application.
Acknowledgments
The authors thank the staff of the Sleep Center of the Affiliated Hospital of Yangzhou University for their support in data collection and sleep monitoring assessment. The authors also thank the Department of Neurology of the Affiliated Hospital of Yangzhou University for providing support during data collection.
Funding Statement
This study was supported by the Yangzhou Natural Science Foundation Youth Fund Project (YZ2017111). Cheng Chu was responsible for Funding Acquisition. The funder had no role in the study design, data collection, data analysis or interpretation, manuscript preparation, or the decision to submit the article for publication.
Abbreviations
OSA, obstructive sleep apnea; AHI, apnea-hypopnea index; NIHSS, National Institutes of Health Stroke Scale; mRS, modified Rankin Scale; BMI, body mass index; NLR, neutrophil-to-lymphocyte ratio; REM, rapid eye movement; NREM, non-rapid eye movement; SpO2, peripheral oxygen saturation; PSQI, Pittsburgh Sleep Quality Index; MoCA, Montreal Cognitive Assessment; CRP, C-reactive protein; AASM, American Academy of Sleep Medicine; AHA/ASA, American Heart Association/American Stroke Association; OR, odds ratio; CI, confidence interval; SE, standard error.
Data Sharing Statement
The datasets generated and/or analyzed during the current study are not publicly available because they contain sensitive patient information and are subject to institutional privacy and ethical restrictions. De-identified data may be available from the corresponding author, Cheng Chu, upon reasonable request and subject to approval by the relevant institutional and ethics authorities.
Ethics Approval and Consent to Participate
This study was approved by the Ethics Committee of Yangzhou University (approval no. 2023-YKL09) and was conducted in accordance with the Declaration of Helsinki. Because this was a retrospective study and all data were anonymized, written informed consent had been obtained from all patients or their legally authorized representatives.
Author Contributions
Yan Liu: Methodology, Investigation, Data Curation, Formal Analysis, Visualization, Writing – Original Draft, Writing – Review & Editing.
Tianqi Zou: Investigation, Data Curation, Validation, Writing – Review & Editing. Zhaoyang Jia: Investigation, Data Curation, Validation, Writing – Review & Editing. Yuxin Ye: Investigation, Data Curation, Validation, Writing – Review & Editing. Tianyu Jing: Investigation, Data Curation, Validation, Writing – Review & Editing. Zhiwen Wu: Investigation, Data Curation, Validation, Writing – Review & Editing. Weixiao Lv: Investigation, Data Curation, Validation, Writing – Review & Editing. Tieyu Tang: Methodology, Writing – Review & Editing.
Gang Xu: Methodology, Writing – Review & Editing.
Ning Luo: Methodology, Supervision, Writing – Review & Editing.
Cheng Chu: Conceptualization, Supervision, Project Administration, Funding Acquisition, Writing – Review & Editing.
All authors gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
All Authors report support from Yangzhou Natural Science Foundation Youth Fund Project (YZ2017111) during the conduct of the study, authors declare that they have no further competing interests.
References
- 1.GBD 2021 Stroke Risk Factor Collaborators. Global, regional, and national burden of stroke and its risk factors, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Neurol. 2024;23(10):973–16. doi: 10.1016/S1474-4422(24)00369-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Tu W-J, Wang L-D. Special Writing Group of China Stroke Surveillance Report. China stroke surveillance report 2021. Mil Med Res. 2023;10(1):33. doi: 10.1186/s40779-023-00463-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lv R, Liu X, Zhang Y, et al. Pathophysiological mechanisms and therapeutic approaches in obstructive sleep apnea syndrome. Signal Transduct Target Ther. 2023;8(1):218. doi: 10.1038/s41392-023-01496-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dharmakulaseelan L, Boulos MI. Sleep apnea and stroke: a narrative review. Chest. 2024;166(4):857–866. doi: 10.1016/j.chest.2024.04.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Su X, Liu S, Wang C, et al. Prevalence, incidence, and the time trends of sleep-disordered breathing among patients with stroke: a systematic review and meta-analysis. Front Neurol. 2024;15:1432085. doi: 10.3389/fneur.2024.1432085 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kurra N, Gandrakota N, Ramakrishnan M, et al. The influence of obstructive sleep apnea on post-stroke complications: a systematic review and meta-analysis. J Clin Med. 2024;13(18):5646. doi: 10.3390/jcm13185646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kleindorfer DO, Towfighi A, Chaturvedi S, et al. 2021 guideline for the prevention of stroke in patients with stroke and transient ischemic attack: a guideline from the American Heart Association/American Stroke Association. Stroke. 2021;52(7):e364–e467. doi: 10.1161/STR.0000000000000375 [DOI] [PubMed] [Google Scholar]
- 8.Intercollegiate Stroke Working Party. National Clinical Guideline for Stroke for the UK and Ireland. 2023. Royal College of Physicians; 2023. [Google Scholar]
- 9.Yang H, Lu S, Yang L. Clinical prediction models for the early diagnosis of obstructive sleep apnea in stroke patients: a systematic review. Syst Rev. 2024;13(1):38. doi: 10.1186/s13643-024-02449-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pivetta B, Chen L, Nagappa M, et al. Use and performance of the STOP-Bang questionnaire for obstructive sleep apnea screening across geographic regions: a systematic review and meta-analysis. JAMA Netw Open. 2021;4(3):e211009. doi: 10.1001/jamanetworkopen.2021.1009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hwang M, Zhang K, Nagappa M, et al. Validation of the STOP-Bang questionnaire as a screening tool for obstructive sleep apnoea in patients with cardiovascular risk factors: a systematic review and meta-analysis. BMJ Open Respir Res. 2021;8(1):e000848. doi: 10.1136/bmjresp-2020-000848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Rha MS, Kim CH, Yoon JH, Cho HJ. Association between the neutrophil-to-lymphocyte ratio and obstructive sleep apnea: a meta-analysis. Sci Rep. 2020;10(1):10862. doi: 10.1038/s41598-020-67708-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med. 2007;4(10):e296. doi: 10.1371/journal.pmed.0040296 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liu L, Li Z, Zhou H, et al. Chinese Stroke Association guidelines for clinical management of ischaemic cerebrovascular diseases: executive summary and 2023 update. Stroke Vasc Neurol. 2023;8(6):e3. doi: 10.1136/svn-2023-002998 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kapur VK, Auckley DH, Chowdhuri S, et al. Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(3):479–504. doi: 10.5664/jcsm.6506 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Troester MM, Quan SF, Berry RB. American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Version 3. American Academy of Sleep Medicine; 2023. [Google Scholar]
- 17.Chung F, Yegneswaran B, Liao P, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108(5):812–821. doi: 10.1097/ALN.0b013e31816d83e4 [DOI] [PubMed] [Google Scholar]
- 18.Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–706. doi: 10.1093/aje/kwh090 [DOI] [PubMed] [Google Scholar]
- 19.Boulos MI, Colelli DR, Vaccarino SR, Kamra M, Murray BJ, Swartz RH. Using a modified version of the STOP-BANG questionnaire and nocturnal oxygen desaturation to predict obstructive sleep apnea after stroke or TIA. Sleep Med. 2019;56:177–183. doi: 10.1016/j.sleep.2018.12.021 [DOI] [PubMed] [Google Scholar]
- 20.Senaratna CV, Perret JL, Lodge CJ, et al. Prevalence of obstructive sleep apnea in the general population: a systematic review. Sleep Med Rev. 2017;34:70–81. doi: 10.1016/j.smrv.2016.07.002 [DOI] [PubMed] [Google Scholar]
- 21.Young T, Skatrud J, Peppard PE. Risk factors for obstructive sleep apnea in adults. JAMA. 2004;291(16):2013–2016. doi: 10.1001/jama.291.16.2013 [DOI] [PubMed] [Google Scholar]
- 22.Young T, Shahar E, Nieto FJ, et al. Predictors of sleep-disordered breathing in community-dwelling adults: the sleep heart health study. Arch Intern Med. 2002;162(8):893–900. doi: 10.1001/archinte.162.8.893 [DOI] [PubMed] [Google Scholar]
- 23.Messineo L, Bakker JP, Cronin J, Yee J, White DP. Obstructive sleep apnea and obesity: a review of epidemiology, pathophysiology and the effect of weight-loss treatments. Sleep Med Rev. 2024;78:101996. doi: 10.1016/j.smrv.2024.101996 [DOI] [PubMed] [Google Scholar]
- 24.Tai JE, Phillips CL, Yee BJ, Grunstein RR. Obstructive sleep apnoea in obesity: a review. Clin Obes. 2024;14(3):e12651. doi: 10.1111/cob.12651 [DOI] [PubMed] [Google Scholar]
- 25.Wang SH, Keenan BT, Wiemken A, et al. Effect of weight loss on upper airway anatomy and the apnea-hypopnea index: the importance of tongue fat. Am J Respir Crit Care Med. 2020;201(6):718–727. doi: 10.1164/rccm.201903-0692OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Javaheri S, Javaheri S, Somers VK, et al. Interactions of obstructive sleep apnea with the pathophysiology of cardiovascular disease, Part 1: JACC State-of-the-Art Review. J Am Coll Cardiol. 2024;84(13):1208–1223. doi: 10.1016/j.jacc.2024.02.059 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wang L, Song Q, Wang C, et al. Neutrophil to lymphocyte ratio predicts poor outcomes after acute ischemic stroke: a cohort study and systematic review. J Neurol Sci. 2019;406:116445. doi: 10.1016/j.jns.2019.116445 [DOI] [PubMed] [Google Scholar]
- 28.Li W, Hou M, Ding Z, Liu X, Shao Y, Li X. Prognostic value of neutrophil-to-lymphocyte ratio in stroke: a systematic review and meta-analysis. Front Neurol. 2021;12:686983. doi: 10.3389/fneur.2021.686983 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ying Y, Yu F, Luo Y, et al. Neutrophil-to-lymphocyte ratio as a predictive biomarker for stroke severity and short-term prognosis in acute ischemic stroke with intracranial atherosclerotic stenosis. Front Neurol. 2021;12:705949. doi: 10.3389/fneur.2021.705949 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Nagappa M, Liao P, Wong J, et al. Validation of the STOP-Bang questionnaire as a screening tool for obstructive sleep apnea among different populations: a systematic review and meta-analysis. PLoS One. 2015;10(12):e0143697. doi: 10.1371/journal.pone.0143697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cho T, Yan E, Chung F. The STOP-Bang questionnaire: a narrative review on its utilization in different populations and settings. Sleep Med Rev. 2024;78:102007. doi: 10.1016/j.smrv.2024.102007 [DOI] [PubMed] [Google Scholar]
- 32.Dempsey JA, Veasey SC, Morgan BJ, O’Donnell CP. Pathophysiology of sleep apnea. Physiol Rev. 2010;90(1):47–112. doi: 10.1152/physrev.00043.2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Malhotra A, White DP. Obstructive sleep apnoea. Lancet. 2002;360(9328):237–245. doi: 10.1016/S0140-6736(02)09464-3 [DOI] [PubMed] [Google Scholar]
- 34.Rashid NH, Zaghi S, Scapuccin M, Camacho M, Certal V, Capasso R. The value of oxygen desaturation index for diagnosing obstructive sleep apnea: a systematic review. Laryngoscope. 2021;131(2):440–447. doi: 10.1002/lary.28663 [DOI] [PubMed] [Google Scholar]
- 35.Bravata DM, Perkins AJ, Myers LJ, et al. Quality improvement intervention to increase sleep apnea diagnostic testing after stroke and transient ischemic attack: a cluster randomized trial. JAMA Netw Open. 2025;8(11):e2543385. doi: 10.1001/jamanetworkopen.2025.43385 [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.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available because they contain sensitive patient information and are subject to institutional privacy and ethical restrictions. De-identified data may be available from the corresponding author, Cheng Chu, upon reasonable request and subject to approval by the relevant institutional and ethics authorities.