Submitted:
29 April 2026
Posted:
30 April 2026
You are already at the latest version
Abstract
Background and Objectives: We aimed to analyze the screening accuracy of the STOP and STOP-Bang questionnaires within three distinct populations from the Mediterranean region: Croatia, Greece, and Türkiye. Additionally, we aimed to optimize previously suggested and to establish population-specific cut-off points for body mass index (BMI) and neck circumference (NC) in the questionnaires to enhance their screening accuracy.
Materials and Methods: A total of 9,102 patients who underwent polysomnography or polygraphy to evaluate suspected OSA were enrolled from: Split Sleep Medicine Centre (Croatia), Ege University Faculty of Medicine (Türkiye), and Thessaloniki G Papaniko-laou Hospital Aristotle University (Greece). Patients completed the STOP and STOP-Bang questionnaires before sleep assessments. Sensitivity, specificity, and area under the curve (AUC) were calculated to assess the screening properties. Additionally, optimized cut-offs for age, NC, and BMI were determined.
Results: The highest AUC values were observed using the STOP-Bang≥5 method, with AUC values of 0.712 for detecting any OSA (AHI≥5/h), 0.684 for moderate or severe OSA (AHI≥15/h), and 0.663 for severe OSA (AHI≥30/h). For individual centers, the STOP-Bang≥5 method performed best in Split, while the STOP≥2+NC method yielded the highest AUCs in Izmir and Thessaloniki for moderate and severe OSA. Optimized cut-off values for age, NC, and BMI improved sensitivity and specificity across all cen-ters.
Conclusions: This study highlights the necessity of population-specific considerations in the screening of OSA. Significant differences in demographics, anthropometrics, symptoms, and comorbidities across populations could impact the questionnaire's screening accuracy. Adjusting age, NC, and BMI cut-off points optimizes the STOP-Bang questionnaire.
Keywords:
obstructive sleep apnea
; STOP
; STOP-Bang
; questionnaire
; screening
1. Introduction
Obstructive sleep apnea (OSA) is an important public health concern due to its increasing prevalence and the frequent appearance of comorbidities [1]. Considering the limited accessibility of polysomnography in certified sleep medicine centers and the increasing workload of professionals and educated experts, screening tests are frequently used for OSA risk assessment. Patients recognized as having a high risk should be further evaluated with diagnostic procedures to objectively determine the presence of OSA. Some of the frequently used screening tools identifying the risk for OSA are the STOP and STOP-Bang questionnaires [2,3]. They both assess symptoms (snoring, tiredness, observed breathing cessations, and hypertension), while STOP-Bang also includes body mass index (BMI), age, sex, and neck circumference (NC) values. STOP-Bang is considered more comprehensive due to a detailed assessment of anthropometric and demographic data. However, previous studies have shown that it has higher sensitivity, but lower specificity compared to the STOP questionnaire, leading to more false-positive results.
Dimensions and composition of the human body, including height, weight, waist-to-hip ratio, NC, and BMI, are of interest since they influence the risk for OSA. More specifically, with increasing BMI, there is an increase in risk for OSA due to a correlation between obesity and airway obstruction during sleep. Additionally, the upper airway and craniofacial features, as well as muscle function, can increase the likelihood of airway collapse during sleep. This is further accentuated by the involvement of NC in risk assessment since excessive fat in the neck region may further narrow the airway during sleep, thus promoting OSA [4]. These anthropometric characteristics can vary significantly among different populations, due to the interplay between numerous factors influencing the anatomy and physiology of the individual and predisposing to OSA [5]. Genetic factors can interact with environmental influences and lifestyle habits, leading to obesity and an increased OSA risk.
Mediterranean populations share certain similarities that indicate common inherited traits, reflecting historical migrations that have shaped these populations. In addition to genetic similarities, these people also share certain lifestyle habits and customs, particularly the Mediterranean diet, which is associated with numerous health benefits and often considered a key factor in longevity [6,7]. However, although quite similar, we believe that even subtle differences among these populations might influence the screening accuracy of the tests relying on anthropometric data.
Previous studies have shown that different demographic and anthropometric characteristics can impact the results and the likelihood of a positive score on the screening questionnaires [8,9]. Thus, adjusting the interpretation of demographic and anthropometric data, as well as the scoring methods of the questionnaires, could enhance their sensitivity and specificity among various patient populations, leading to better identification and selection of patients at risk for OSA [10].
Thus, this study aimed to comprehensively analyze the screening accuracy regarding sensitivity and specificity of the STOP and STOP-Bang questionnaires within three distinct populations from the Mediterranean region: Croatia, Greece, and Türkiye. Additionally, our secondary objective was to optimize the suggested and establish population-specific cut-off points for BMI and NC in questionnaires to enhance their screening accuracy.
2. Materials and Methods
2.1. Study Population
This study was approved by the Biomedical Research Ethics Committee of the University of Split, School of Medicine for Croatia (approval number 2181-198-03-04-14-0027), Institutional Ethics Committee and Internal Review Board of Faculty of Medicine, Ege University for Türkiye (approval number 12-1.1/7), and for Thessaloniki (approval number 357/20160323). All study procedures were performed according to the ethical principles of the 1964 Declaration of Helsinki and its later amendments. All patients before the sleep assessment signed informed consent for participation.
The study was performed in the Split Sleep Medicine Centre (Split, Croatia), Ege University Faculty of Medicine Department of Respiratory Medicine (Izmir, Türkiye), and Aristotle University G Papanikolaou Hospital (Thessaloniki, Greece) during 2012-2023. A total of 9,102 patients who underwent full-night polysomnography (PSG) or polygraphy (PG) to evaluate suspected OSA during the study period were enrolled. In Split, 1,023 (32.9%) patients were assessed with PSG and 2,085 (67.1%) with PG. In Izmir, a total of 3,637 patients underwent PSG, while in Thessaloniki, all (2,357) patients were evaluated with PG.
2.2. Polysomnography
Full-night PSG/PG was done and scored, and OSA was diagnosed according to the criteria of the American Academy of Sleep Medicine (AASM) Manual for Scoring Sleep and Associated Events and the European Sleep Research Society (ESRS) guidelines. In Split, Croatia, for full-night PG recording, either SOMNOcheck2 (Weinmann, Hamburg, Germany) or Alice Night One (Philips Respironics, Eindhoven, the Netherlands) was used, while for PSG, Alice 6 (Philips Respironics, Eindhoven, the Netherlands) was used. In Izmir, Türkiye, Alice 5 (Philips, Respironics, USA) was used. In Thessaloniki, Greece, either SOMNOcheck2 (Weinmann, Hamburg, Germany) or Nox T3 (Nox medical, Iceland) was used. Apnea-hypopnea index (AHI) <5 events/h was considered as no OSA, and data were grouped according to the established severity levels of the baseline AHI (mild ≥5 to <15 events/h, moderate ≥15 to <30 events/h and severe ≥30 events/h).
2.3. STOP and STOP-Bang Questionnaires
On the night of admission, before the sleep assessment, all patients filled out the STOP and STOP-Bang questionnaires. Both the STOP and the STOP-Bang questionnaires are easy-to-use tools to screen patients for OSA and have good reliability and internal consistency. The STOP questionnaire consists of 4 anamnestic questions: snoring (S), tiredness (T), observed breathing cessations (O), and arterial pressure (P); while Bang considers anthropometric and demographic data such as BMI (B), age (A), neck circumference (N), and gender (G). The STOP questionnaire indicates high OSA risk in case of two or more positive answers. However, there are various methods to interpret the STOP-Bang questionnaire: it is considered positive when the STOP (2 or more positive answers) is positive plus one additional component of the Bang section, or when there are 3 or more, or 5 or more positive answers in the STOP-Bang questionnaire [2,3].
2.4. Statistical Analysis
Data were analyzed using the Statistical Package for the Social Sciences (SPSS, version 20.0 for Windows) program. Continuous variables were reported as means±standard deviations, while categorical variables as counts (percentages). Before analysis, the normality of data was assessed via the Shapiro-Wilk test.
One-way analysis of variance (ANOVA) was used to compare demographic, anthropometric, and sleep parameter data among three centers, followed by Tukey’s post hoc test to identify pairwise differences. Differences in symptom prevalence, comorbidities, and other categorical clinical characteristics were evaluated using the chi-square test. For the evaluation of screening performance and accuracy of the STOP and STOP-Bang questionnaires, receiver operating characteristic (ROC) curve analyses were used. Area under the curve (AUC) values were calculated for detecting OSA at various severity thresholds (e.g., AHI≥5/h, AHI≥15/h, and AHI≥30/h). Additionally, sensitivity and specificity values were reported. The optimal cut-off values for age, NC, and BMI were determined using the Youden index to maximize the combined sensitivity and specificity.
All statistical tests were two-tailed, and a p<0.05 was considered statistically significant.
3. Results
3.1. Demographic and Anthropometric Data
A total of 9,102 patients (69.5% male) from Greece, Türkiye, and Croatia were included. Patients from Thessaloniki, Greece (2,357; 25.9%) underwent full-night PG, while those from Izmir, Türkiye (3,637; 40%) and Split, Croatia (3,108; 34.1%) were assessed using PG (67.1%) or PSG (32.9%). There were no differences in sex distribution among centers. Thessaloniki patients had the highest body weight (98.8±22.5 kg) and BMI (33.4±7.1 kg/m²), while Split patients were the tallest (1.77±0.10 m) (all p<0.001). Demographic and anthropometric data are presented in Table 1.
Sex data are shown as absolute and relative frequencies, all other data are shown as means ± standard deviations.
3.2. Clinical Characteristics
Izmir patients most frequently reported snoring (98.2%), daytime tiredness (79.5%), and observed breathing cessations (81.4%), while Split patients had the highest prevalence of arterial hypertension (45.9%). Diabetes, depression, and gastroesophageal reflux disease were most common in Izmir (Table 2).
3.3. Sleep Parameters
The average AHI was highest among patients from Izmir (37.7±28.8/h), compared to those from Thessaloniki (32.7±24.8/h) and Split (25.3±22.4/h) (p<0.001). The mean blood oxygen saturation of patients from Thessaloniki (91.9±3.4%) was lower than that of the patients from Izmir (92.7±14.4%) and Split (94.2±3.3%) (p<0.001) (Table 2).
3.4. Screening Performance of STOP and STOP-Bang Questionnaires Across Different Populations
Table 3 presents the sensitivity, specificity, and AUC values of the STOP and different STOP-Bang assessment methods. The STOP-Bang≥5 method had the best screening properties overall (AUC for OSA: 0.712, for moderate or severe OSA: 0.684, for severe OSA: 0.663).
In Thessaloniki, STOP-Bang≥3 was most effective for OSA, while STOP≥2+NC was best for moderate to severe OSA. In Izmir, STOP-Bang≥5 was optimal for detecting OSA and moderate or severe OSA, while STOP≥2+NC was best for severe OSA. In Split, STOP-Bang≥5 was the best for all categories (AUCs: 0.720 for OSA, 0.707 for moderate/severe OSA, 0.701 for severe OSA).
3.5. Impact of Optimized Cut-Offs on Screening Accuracy of the STOP-Bang Questionnaire
Cut-off points for age, NC, and BMI were optimized to improve the screening properties of the STOP-Bang questionnaire (Table 4). Age cut-offs ranged from 45.5 years (Split) to 54.5 years (Izmir), NC ranged from 40.5 cm (Thessaloniki) to 40.8 cm (Izmir and Split), and BMI ranged from 26.4 kg/m² (Split) to 30.7 kg/m² (Thessaloniki).
Finally, when these optimized cut-off values were applied, changes in the screening accuracy of the STOP-Bang questionnaire were further increased sensitivity and decreased specificity, and reflected in the changes in AUC values of the STOP-Bang questionnaire in each population (Table 5).
4. Discussion
A large cohort of 9,102 patients presenting to sleep medicine centers for OSA diagnostics from three Mediterranean countries, Greece, Türkiye, and Croatia, was included in this study. The studied populations share many similarities, possibly due to their common Mediterranean background, reflected in lifestyle and genetic traits associated with various health benefits, such as cardiovascular health and longevity. Populations with OSA in Mediterranean countries, such as Greece, Türkiye, and Croatia, differ from other OSA populations due to specific demographic, anthropometric, and health factors. However, despite these common characteristics, notable differences among these populations might have influenced the risk factors, presentation, and OSA severity. These differences likely include variations in anthropometric measures, such as obesity and BMI, as well as the prevalence of comorbidities like hypertension, diabetes, and depression. Our study revealed that patients in Greece had a higher BMI, which is associated with more severe OSA, while higher rates of comorbidities like diabetes and GERD were observed in Türkiye. These factors appeared to influence both the presentation and severity of OSA. One might argue on how much these differences might affect the results of OSA screening tests, specifically having in mind that it has been previously proposed that inconsistencies in epidemiological studies might come as a consequence of racial and/or ethnic profile, ethnicity, study design, age, BMI and sex [11]. It is important to note that the optimal cut-off points for risk assessment using the STOP-Bang questionnaire vary across populations, highlighting the need to tailor diagnostic methods to the specific characteristics of each region. These variations suggest that population-specific adaptations in screening and diagnosis are essential for more accurate OSA detection. Our findings emphasize the importance of considering these differences when applying screening tools, particularly the STOP-Bang questionnaire. Notably, our study showed that different assessment methods for the STOP-Bang questionnaire resulted in varying screening accuracy across the three populations, with each region’s distinct anthropometric characteristics influencing the optimal cut-off points.
In terms of OSA screening, the STOP and STOP-Bang questionnaires are reliable and widely used tools for OSA risk assessment [12,13,14]. Still, the assessment method for interpreting the STOP-Bang questionnaire resulting in the best screening accuracy in different populations is questionable [15,16]. Results from the present study revealed that the best screening properties with the highest AUC values for detecting mild, moderate, and severe OSA had STOP-Bang≥5 method in the total sample, but not when three populations were analyzed separately. The differences in predictive values, such as the lower specificity of the STOP-Bang questionnaire in Thessaloniki and Izmir compared to Split, can be attributed to several factors. First, the anthropometric characteristics of the populations vary, with patients in Thessaloniki and Izmir having higher BMI and NC, which could lead to higher rates of positive screenings even in individuals without severe OSA. Additionally, differences in the prevalence of comorbidities such as hypertension may also influence the specificity of the tool. The variations in the cut-off points used for the STOP-Bang questionnaire in each center further reflect these population differences, indicating that regional factors influence the tool’s performance. Thus, the specific characteristics of each population, including anthropometric and demographic factors, and comorbidities likely contribute to the observed discrepancies in predictive values. In Thessaloniki, the STOP-Bang≥3 method was most effective for detecting OSA, while the STOP≥2+NC method performed best for detecting moderate and severe patients. The same method performed the best to detect patients with severe OSA in Izmir, while STOP-Bang≥5 method was better in screening patients with moderate OSA. In Split, STOP-Bang≥5 method was the best assessment method, in all OSA severity groups. These results suggest that using different assessment methods of the STOP-Bang questionnaire in different populations is important to enhance screening accuracy, especially when considering anthropometric differences between populations. Our findings support the idea that although STOP-Bang is a reliable tool, its interpretation method needs to be adjusted to optimize its sensitivity and specificity in different populations.
Although all studied populations belonged to the Mediterranean region and shared some genetic and lifestyle traits, differences in anthropometric measures such as body weight, height, and BMI were still evident across three populations, possibly reflecting different OSA severity reported in included patients. Patients from Thessaloniki exhibited the highest body weight and BMI, factors known to contribute to the risk and severity of OSA [17,18]. In particular, the assessment of OSA risk might be influenced by the differing prevalence of obesity, a well-established risk factor for OSA, across these cohorts [11]. The emerging threats of obesity, along with an aging population, significantly contribute to the global burden of OSA, and impact the health of entire populations. With the persistent obesity epidemic, a rise in prevalence is imminent in countries related to the Mediterranean region since recent evidence supports an increased obesity and metabolic syndrome in OSA patients from Mediterranean countries such as Greece, Spain, and Italy [19].
Furthermore, when considering the application of the STOP-Bang questionnaire across different geographic populations concerning varying anthropometry, it is essential to recognize that anthropometric factors, such as body weight, NC, and others, may exhibit variations among diverse ethnic groups [20,21,22]. This is particularly crucial since some of these factors are integral to the questionnaire itself. Thus, to further improve the STOP-Bang questionnaire's screening accuracy, the cut-off points for age, NC, and BMI were optimized in each studied population. The optimal cut-off values, which varied between populations, reflected demographic and anthropometric differences. For instance, Thessaloniki had the highest optimal age cut-off, whereas Split had the lowest, suggesting that OSA risk in these populations increases at different ages. Similarly, NC cut-off points were slightly higher in Izmir and Split compared to Thessaloniki, highlighting the importance of tailoring screening criteria to population-specific anthropometric characteristics. The optimized BMI cut-off values were also lower in Split, consistent with the lower average BMI in this population, further supporting the need for regional adaptations of screening tools to enhance their accuracy. One might conclude that it is important to adjust risk thresholds to reflect the anthropometric characteristics specific to a particular geographic population.
In the present study, high rates of upper airway symptoms such as snoring and observed breathing cessations combined with daytime tiredness were reported in all populations studied. Patients from Izmir had the highest average AHI values and highest rates of reported symptoms, emphasizing the association between clinical symptoms and OSA severity. It has been reported that OSA is associated with many comorbid disorders affecting the cardiovascular, respiratory, metabolic, and central nervous system, leading to adverse health outcomes [11]. The higher prevalence of comorbidities such as diabetes, asthma, and GERD in Izmir further supports the notion that the OSA severity might be associated with unfavorable health outcomes [23,24]. However, despite having the lowest average AHI, patients from Split reported the highest prevalence of comorbid arterial hypertension, an established risk factor for cardiovascular complications in OSA patients [25]. The results of our study indicate that even though all studied populations belong to the Mediterranean region, some inter-ethnic factors may contribute to observed differences in susceptibility to comorbidities and be reflected in the accuracy of screening properties of the STOP-Bang questionnaire. Thus, further optimization and tailoring of the screening interpretation may help to better identify patients with a higher OSA risk requiring prompt care [11].
The main strength of our study is that it is one of the largest conducted in university hospitals across three countries, where demographic and anthropometric characteristics, detailed medical history, and sleep test (PSG/PG) data were available for all patients. However, the inclusion of patients from three different sleep centers, located in distinct geographic regions, introduces potential limitations and biases in terms of the patient populations studied, patient selection, and clinical protocols used. We acknowledge that the patients included in our study were diagnosed using two different methods, PG and PSG. However, both are valid diagnostic tools for cases with a high clinical suspicion of uncomplicated OSA. The choice between these methods depends on the clinical setting and available resources. Additionally, since the study was conducted on patients from sleep clinics, there is an over-representation of OSA patients, which limits the generalizability of the results to the broader population.
5. Conclusions
In conclusion, these findings suggest that the one-fits-all approach to OSA screening is not optimal, emphasizing the need for population-specific considerations, even when they share common lifestyle factors. Significant differences in demographics, anthropometric measures, symptom prevalence, and comorbidities across populations might impact the accuracy of screening tools, such as the STOP-Bang questionnaire. Therefore, it is crucial to adapt the interpretation of this questionnaire to fit the unique characteristics of each population, including anthropometrics, obesity rates, or comorbidities. Finally, early detection of high-risk patients is essential not only for improving individual health outcomes but also for reducing the economic burden on national healthcare systems, leading to better long-term outcomes and more cost-effective management of OSA across different populations.
Author Contributions
Conceptualization, I.P.D., R.P., N.I., L.L.K., O.K.B., A.P., M.S.T., S.K. and Z.D.; methodology, N.I., O.K.B., A.P., M.S.T, S.K. and Z.D.; formal analysis, L.L.K.; investigation, I.P.D., R.P., O.K.B., A.P. and Z.D.; data curation, I.P.D., N.I., M.S.T. and S.K.; writing—original draft preparation, I.P.D., R.P., L.L.K., Z.D.; writing—review and editing, I.P.D., R.P., N.I., O.K.B., A.P., M.S.T., S.K. and Z.D.; visualization, L.L.K., I.P.D., R.P., O.K.B., A.P., Z.D.; supervision R.P., O.K.B., A.P. and Z.D. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Institutional Review Board Statement
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Biomedical Research Ethics Committee of the University of Split, School of Medicine for Croatia (approval number 2181-198-03-04-14-0027), Institutional Ethics Committee and Internal Review Board of Faculty of Medicine, Ege University for Türkiye (approval number 12-1.1/7), and for Thessaloniki (approval number 357/20160323).
Informed Consent Statement
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement
All relevant data are included in the manuscript; raw data are available from the corresponding author upon reasonable request.
Acknowledgments
During the preparation of this manuscript, the authors used ChatGPT (OpenAI), models GPT-4 and GPT-5 for the purposes of refining and editing the English language of this manuscript, particularly in improving clarity, grammar, and phrasing. The authors have reviewed and edited the output and take full responsibility for the content of this publication.
This study was not directly conducted within the ESADA framework; however, the authors acknowledge the European Sleep Apnea Database (ESADA) network for facilitating inter-centre collaboration relevant to this work.
Conflicts of Interest
The authors declare no conflicts of interest.
Abbreviations
The following abbreviations are used in this manuscript:
| OSA | Obstructive sleep apnea |
| BMI | Body mass index |
| NC | Neck circumference |
| PSG | Polysomnography |
| PG | Polygraphy |
| AHI | Apnea-Hypopnea Index |
| GERD | Gastroesophageal reflux disease |
| AUC | Area under the curve |
| ROC | Receiver operating characteristic |
| ANOVA | Analysis of variance |
References
- Faria, A.; Allen, A.H.; Fox, N.; Ayas, N.; Laher, I. The public health burden of obstructive sleep apnea. Sleep. Sci. 2021, 14, 257–65. [Google Scholar] [CrossRef]
- Chung, F.; Yegneswaran, B.; Liao, P.; Chung, S.A.; Vairavanathan, S.; Islam, S. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology 2008, 108, 812–21. [Google Scholar] [CrossRef] [PubMed]
- Chung, F.; Abdullah, H.R.; Liao, P. STOP-Bang Questionnaire: A practical approach to screen for obstructive sleep apnea. Chest 2016, 149, 631–8. [Google Scholar] [CrossRef] [PubMed]
- Patel, H.; Chapman, T.; Whitehead, R.; Zhang, Y.; Ritz, E.; Mokhlesi, B. The Impact of Neck Circumference and BMI on Upper Airway Collapsibility and Risk of Complete Concentric Collapse in OSA. Laryngoscope Epub Ahead Print. 2025. [Google Scholar] [CrossRef]
- Goyal, A.; Wachinou, A.P.; Imler, T.; Solelhac, G.; Joshi, A.; Agodokpessi, G. Obesity, obstructive sleep Apnea, and cardiometabolic risk: A comparative analysis of population-Based cohorts in Switzerland, India, and Benin. Sleep. Breath. 2025, 29, 346. [Google Scholar] [CrossRef]
- Dominguez, L.J.; Di Bella, G.; Veronese, N.; Barbagallo, M. Impact of Mediterranean diet on chronic non-communicable diseases and longevity. Nutrients 2021, 13, 2028. [Google Scholar] [CrossRef]
- Mazza, E.; Ferro, Y.; Pujia, R.; Mare, R.; Maurotti, S.; Montalcini, T. Mediterranean diet in healthy aging. J. Nutr. Health Aging 2021, 25, 1076–83. [Google Scholar] [CrossRef]
- Waseem, R.; Chan, M.T.V.; Wang, C.Y.; Seet, E.; Tam, S.; Loo, S.Y. Diagnostic performance of the STOP-Bang questionnaire as a screening tool for obstructive sleep apnea in different ethnic groups. J. Clin. Sleep. Med. 2021, 17, 521–32. [Google Scholar] [CrossRef]
- 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. [Google Scholar] [CrossRef]
- Seguin, L.; Tamisier, R.; Deletombe, B.; Lopez, M.; Pepin, J.L.; Payen, J.F. Preoperative screening for obstructive sleep apnea using alternative scoring models of the Sleep Tiredness Observed Pressure-Body Mass Index Age Neck Circumference Gender Questionnaire: An External Validation. AnesthAnalg 2020, 131, 1025–31. [Google Scholar] [CrossRef] [PubMed]
- Lyons, M.M.; Bhatt, N.Y.; Pack, A.I.; Magalang, U.J. Global burden of sleep-disordered breathing and its implications. Respirology 2020, 25, 690–702. [Google Scholar] [CrossRef] [PubMed]
- Nagappa, M.; Liao, P.; Wong, J.; Auckley, D.; Ramachandran, S.K.; Memtsoudis, S. 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, e0143697. [Google Scholar] [CrossRef] [PubMed]
- Chiu, H.Y.; Chen, P.Y.; Chuang, L.P.; Chen, N.H.; Tu, Y.K.; Hsieh, Y.J. Diagnostic accuracy of the Berlin questionnaire, STOP-BANG, STOP, and Epworth Sleepiness Scale in detecting obstructive sleep apnea: A bivariate meta-analysis. Sleep. Med. Rev. 2017, 36, 57–70. [Google Scholar] [CrossRef]
- Duarte, R.L.M.; Magalhães-da-Silveira, F.J.; Gozal, D. Screening for obstructive sleep apnea: comparing the American Academy of Sleep Medicine proposed criteria with the STOP-Bang, NoSAS, and GOAL instruments. J. Clin. Sleep. Med. 2023, 19, 1239–46. [Google Scholar] [CrossRef]
- Neves Junior, J.A.S.; Fernandes, A.P.A.; Tardelli, M.A.; Yamashita, A.M.; Moura, S.M.P.G.T.; Tufik, S.; Silva, H.C.A.D. Cutoff points in STOP-Bang questionnaire for obstructive sleep apnea. ArqNeuropsiquiatr 2020, 78, 561–9. [Google Scholar] [CrossRef]
- Nagappa, M.; Wong, J.; Singh, M.; Wong, D.T.; Chung, F. An update on the various practical applications of the STOP-Bang questionnaire in anesthesia, surgery, and perioperative medicine. Curr. OpinAnaesthesiol 2017, 30, 118–25. [Google Scholar] [CrossRef]
- Drager, L.F.; Togeiro, S.M.; Polotsky, V.Y.; Lorenzi-Filho, G. Obstructive sleep apnea: a cardiometabolic risk in obesity and the metabolic syndrome. J. Am. Coll. Cardiol. 2013, 62, 569–76. [Google Scholar] [CrossRef] [PubMed]
- Alenezi, M.A.; Alabdulathim, S.; Alhejaili, S.A.M.; Al Sheif, Z.A.A.; Aldossari, K.K.; Bakhsh, J.I. The association between obesity and the development and severity of obstructive sleep apnea: A systematic review. Cureus 2024, 16, e69962. [Google Scholar] [CrossRef]
- Zito, A.; Steiropoulos, P.; Barceló, A.; Marrone, O.; Esquinas, C.; Buttacavoli, M. Obstructive sleep apnoea and metabolic syndrome in Mediterranean countries. Eur. Respir. J. 2011, 37, 717–9. [Google Scholar] [CrossRef]
- Patel, D.; Tsang, J.; Saripella, A.; Nagappa, M.; Islam, S.; Englesakis, M. Validation of the STOP questionnaire as a screening tool for OSA among different populations: a systematic review and meta-regression analysis. J. Clin. Sleep. Med. 2022, 18, 1441–53. [Google Scholar] [CrossRef]
- Pivetta, B.; Chen, L.; Nagappa, M.; Saripella, A.; Waseem, R.; Englesakis, M. 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, e211009. [Google Scholar] [CrossRef]
- Loh, J.M.-R.; Toh, S.-T. Rethinking neck circumference in STOP-BANG for Asian OSA. Proc. Singap. Healthc. 2019, 28, 105–9. [Google Scholar] [CrossRef]
- Al-Jahdali, H.; Ahmed, A.E.; Abdullah, A.H.; Ayaz, K.; Ahmed, A.; Majed, A. Comorbidities in clinical and polysomnographic features of obstructive sleep apnea: A single tertiary care center experience. J. Epidemiol. Glob. Health 2022, 12, 486–95. [Google Scholar] [CrossRef] [PubMed]
- Stansbury, R.C.; Strollo, P.J. Clinical manifestations of sleep apnea. J. Thorac. Dis. 2015, 7, E298-310. [Google Scholar] [CrossRef]
- Seravalle, G.; Grassi, G. Sleep apnea and hypertension (2022). High. Blood Press Cardiovasc Prev. 29, 23–31. [CrossRef] [PubMed]
Table 1.
Demographic and anthropometric data of the study population.
| Overall N=9,102 |
Thessaloniki Greece N=2,357 (25.9%) |
Izmir Türkiye N=3,637 (40%) |
Split Croatia N=3,108 (34.1%) |
p | |
|---|---|---|---|---|---|
| Age | 53.82±13.03 | 57.22±13.07 | 51.21±12.02 | 54.30±13.47 | <0.001 1 |
| Sex | |||||
| Male, N (%) | 6322 (69.5) | 1683 (71.4) | 2496 (68.6) | 2143 (69) | 0.056 2 |
| Female, N (%) | 2780 (30.5) | 674 (28.6) | 1141 (31.4) | 965 (31) | |
| Weight (kg) | 94.6±20.7 | 98.8±22.5 | 92.4±18.8 | 94.1±20.9 | <0.001 1 |
| Height (m) | 1.72±0.10 | 1.72±0.09 | 1.69±0.10 | 1.77±0.10 | <0.001 1 |
| Body mass index (kg/m2) | 31.9±6.7 | 33.4±7.1 | 32.6±6.8 | 29.8±5.8 | <0.001 1 |
| Neck circumference (cm) | 41.5±4.8 | 41.5±5.3 | 41.4±4.0 | 41.5±5.2 | 0.842 1 |
1 One-way ANOVA; 2 Chi-square test
Table 2.
Clinical characteristics and sleep parameters of the study population.
| Overall N=9,102 |
Thessaloniki Greece N=2,357 (25.9%) | Izmir Türkiye N=3,637 (40%) |
Split Croatia N=3,108 (34.1%) |
p 1 | ||
|---|---|---|---|---|---|---|
| Clinical characteristics | ||||||
| Snoring | ||||||
| Yes, N (%) | 8,109 (89.7) | 2,247 (95.3) | 3,573 (98.2) | 2,289 (75.2) | <0.001 | |
| Tiredness | ||||||
| Yes, N (%) | 6,778 (74.8) | 1,588 (67.4) | 2,891 (79.5) | 2,299 (75.0) | <0.001 | |
| Observed breathing cessations | ||||||
| Yes, N (%) | 6,661 (73.9) | 1,783 (75.6) | 2,960 (81.4) | 1,918 (63.5) | <0.001 | |
| Hypertension | ||||||
| Yes, N (%) | 3,863 (42.4) | 980 (41.6) | 1,455 (40.0) | 1,428 (45.9) | <0.001 | |
| Diabetes mellitus | ||||||
| Yes, N (%) | 1,740 (19.2) | 382 (16.2) | 893 (24.6) | 465 (15.1) | <0.001 | |
| Asthma | ||||||
| Yes, N (%) | 1,382 (15.2) | 61 (2.6) | 958 (26.3) | 363 (11.8) | <0.001 | |
| Depression | ||||||
| Yes, N (%) | 408 (6.1) | 0 2 | 210 (5.8) | 198 (6.4) | <0.001 | |
| GERD | ||||||
| Yes, N (%) | 2,078 (22.9) | 70 (3) | 1,116 (30.7) | 892 (29.1) | <0.001 | |
| Sleep parameters | ||||||
| Type of sleep test | ||||||
| PSG, N (%) | 4,660 (51.2) | 0 (0) | 3,637 (100) | 1,023 (32.9) | <0.001 | |
| PG, N (%) | 4,442 (48.8) | 2,357 (100) | 0 (0) | 2,085 (67.1) | ||
| PG | PSG | PG | PSG | |||
| OSA severity | ||||||
| No OSA, N (%) | 1,065 (11.7) | 318 (13.5) | 279 (7.7) | 258 (12.4) | 210 (20.6) | <0.001 |
| Mild OSA, N (%) | 1,940 (21.3) | 353 (15) | 679 (18.7) | 651 (31.2) | 257 (25.1) | |
| Moderate OSA, N (%) | 1,996 (21.9) | 566 (24) | 845 (23.2) | 412 (19.8) | 173 (16.9) | |
| Severe OSA, N (%) | 4,101 (45.1) | 1,120 (47.5) | 1,834 (50.4) | 764 (36.6) | 383 (37.4) | |
| AHI (events/h) | 32.2±26.3 | 32.7±24.8 | 37.7±28.8 | 25.3±22.4 | <0.001 | |
| Mean O2 saturation (%) | 93.0±9.5 | 91.9±3.4 | 92.7±14.4 | 94.2±3.3 | <0.001 | |
| Lowest O2 saturation (%) | 79.4±11.2 | 78.9±9.8 | 78.9±12.2 | 80.5±11.0 | <0.001 | |
1 Chi-square test; 2 Missing data. All data are shown as absolute and relative frequencies. Abbreviations: AHI, apnea-hypopnea index; GERD, gastroesophageal reflux disease; OSA, obstructive sleep apnea.
Table 3.
Sensitivity, specificity, and AUC values of STOP and various STOP-Bang assessment methods in the study population.
Table 3.
Sensitivity, specificity, and AUC values of STOP and various STOP-Bang assessment methods in the study population.
| AHI | Overall | Thessaloniki, Greece | Izmir, Türkiye | Split, Croatia | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ≥5 | ≥15 | ≥30 | ≥5 | ≥15 | ≥30 | ≥5 | ≥15 | ≥30 | ≥5 | ≥15 | ≥30 | ||||||
| STOP≥2 | |||||||||||||||||
| Sensitivity | 92.7 | 94.9 | 96.5 | 93.0 | 94.7 | 96.1 | 97.4 | 97.9 | 98.8 | 86.5 | 90.6 | 93.3 | |||||
| Specificity | 29.1 | 19.5 | 15.0 | 23.6 | 18.9 | 14.0 | 10.0 | 6.0 | 5.1 | 44.6 | 29.3 | 24.8 | |||||
| AUC | 0.609 | 0.572 | 0.558 | 0.583 | 0.568 | 0.550 | 0.537 | 0.520 | 0.520 | 0.655 | 0.599 | 0.591 | |||||
| STOP≥2 + BMI | |||||||||||||||||
| Sensitivity | 23.3 | 27.5 | 32.1 | 37.1 | 41.9 | 49.2 | 30.0 | 32.6 | 37.4 | 4.1 | 5.5 | 6.9 | |||||
| Specificity | 92.1 | 90.6 | 87.1 | 87.1 | 86.4 | 80.1 | 86.0 | 81.7 | 79.9 | 99.3 | 98.9 | 98.4 | |||||
| AUC | 0.577 | 0.590 | 0.596 | 0.621 | 0.641 | 0.646 | 0.580 | 0.571 | 0.587 | 0.517 | 0.522 | 0.526 | |||||
| STOP≥2 + Age | |||||||||||||||||
| Sensitivity | 59.4 | 62.7 | 63.4 | 67.5 | 69.8 | 69.6 | 53.9 | 55.6 | 56.2 | 60.3 | 66.8 | 68.9 | |||||
| Specificity | 69.8 | 57.6 | 50.0 | 59.7 | 51.1 | 41.4 | 73.5 | 59.0 | 52.7 | 74.6 | 59.8 | 53.0 | |||||
| AUC | 0.646 | 0.602 | 0.567 | 0.636 | 0.605 | 0.556 | 0.637 | 0.573 | 0.545 | 0.674 | 0.633 | 0.609 | |||||
| STOP≥2 + Neck | |||||||||||||||||
| Sensitivity | 58.8 | 65.1 | 71.8 | 59.4 | 64.8 | 70.4 | 61.3 | 65.7 | 72.5 | 55.1 | 64.5 | 72.0 | |||||
| Specificity | 79.0 | 67.5 | 59.9 | 80.8 | 73.3 | 60.9 | 67.4 | 59.5 | 54.5 | 84.8 | 70.2 | 64.2 | |||||
| AUC | 0.689 | 0.663 | 0.658 | 0.701 | 0.692 | 0.657 | 0.643 | 0.626 | 0.635 | 0.700 | 0.674 | 0.681 | |||||
| STOP≥2 + Sex | |||||||||||||||||
| Sensitivity | 66.6 | 71.1 | 75.2 | 68.2 | 70.5 | 73.7 | 68.2 | 71.1 | 74.7 | 63.2 | 71.6 | 77.4 | |||||
| Specificity | 64.0 | 53.4 | 47.0 | 56.9 | 49.5 | 43.2 | 54.5 | 46.5 | 41.8 | 74.6 | 60.2 | 54.1 | |||||
| AUC | 0.652 | 0.622 | 0.611 | 0.626 | 0.600 | 0.584 | 0.613 | 0.588 | 0.582 | 0.689 | 0.659 | 0.657 | |||||
| STOP-Bang≥3 | |||||||||||||||||
| Sensitivity | 96.2 | 98.4 | 99.3 | 97.7 | 98.6 | 99.6 | 98.2 | 98.8 | 99.6 | 92.5 | 97.5 | 98.7 | |||||
| Specificity | 29.3 | 17.2 | 11.7 | 20.4 | 13.3 | 8.6 | 12.9 | 6.6 | 4.9 | 45.4 | 26.5 | 20.0 | |||||
| AUC | 0.628 | 0.578 | 0.555 | 0.721 | 0.560 | 0.541 | 0.556 | 0.527 | 0.522 | 0.690 | 0.620 | 0.594 | |||||
| STOP-Bang≥5 | |||||||||||||||||
| Sensitivity | 67.2 | 74.3 | 80.1 | 72.5 | 77.2 | 81.5 | 70.3 | 74.9 | 80.7 | 59.2 | 70.8 | 77.9 | |||||
| Specificity | 75.1 | 62.3 | 52.4 | 71.7 | 60.1 | 47.0 | 63.1 | 52.2 | 45.4 | 84.8 | 70.5 | 62.2 | |||||
| AUC | 0.712 | 0.684 | 0.663 | 0.591 | 0.687 | 0.643 | 0.667 | 0.636 | 0.630 | 0.720 | 0.707 | 0.701 | |||||
Abbreviations: AUC, area under the curve; AHI, apnea-hypopnea index; BMI, body mass index.
Table 4.
Cut-off points defined based on the maximum level of sensitivity and specificity (Youden index approach) for the detection of OSA (AHI≥5 events/h).
Table 4.
Cut-off points defined based on the maximum level of sensitivity and specificity (Youden index approach) for the detection of OSA (AHI≥5 events/h).
| Cut-off | Sensitivity | Specificity | ||
|---|---|---|---|---|
| Thessaloniki, Greece | Age (years) | 47.21 | 0.789 | 0.437 |
| Izmir, Türkiye | Age (years) | 54.50 | 0.426 | 0.832 |
| Split, Croatia | Age (years) | 45.50 | 0.807 | 0.529 |
| Thessaloniki, Greece | Neck (cm) | 40.50 | 0.631 | 0.764 |
| Izmir, Türkiye | Neck (cm) | 40.75 | 0.620 | 0.670 |
| Split, Croatia | Neck (cm) | 40.75 | 0.642 | 0.771 |
| Thessaloniki, Greece | BMI (kg/m2) | 30.71 | 0.644 | 0.720 |
| Izmir, Türkiye | BMI (kg/m2) | 29.74 | 0.636 | 0.609 |
| Split, Croatia | BMI (kg/m2) | 26.41 | 0.779 | 0.611 |
Abbreviations: AHI, apnea-hypopnea index; BMI, body mass index; OSA, obstructive sleep apnea.
Table 5.
Differences in the sensitivity, specificity, and AUC values of the STOP-Bang questionnaire with the use of previous and newly optimized cut-off points for body mass index, neck circumference, and age in the study population.
Table 5.
Differences in the sensitivity, specificity, and AUC values of the STOP-Bang questionnaire with the use of previous and newly optimized cut-off points for body mass index, neck circumference, and age in the study population.
| Thessaloniki, Greece | Izmir, Türkiye |
Split, Croatia |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AHI | ≥5 | ≥15 | ≥30 | ≥5 | ≥15 | ≥30 | ≥5 | ≥15 | ≥30 | |||
| STOP-Bang ≥3 previous(p) vs. new(n) cut-off points | Sensitivityp | 97.7 | 98.6 | 99.6 | 98.2 | 98.8 | 99.6 | 92.5 | 97.5 | 98.7 | ||
| Specificityp | 20.4 | 13.3 | 8.6 | 12.9 | 6.6 | 4.9 | 45.4 | 26.5 | 20 | |||
| AUCp | 0.721 | 0.560 | 0.541 | 0.556 | 0.527 | 0.522 | 0.690 | 0.620 | 0.594 | |||
| Sensitivityn | 98.6 | 99.2 | 99.7 | 98.2 | 98.9 | 99.6 | 98.2 | 98.9 | 99.6 | |||
| Specificityn | 19.2 | 11.4 | 7.0 | 12.2 | 6.8 | 4.8 | 12.2 | 6.8 | 4.8 | |||
| AUCn | 0.589 | 0.553 | 0.534 | 0.552 | 0.528 | 0.522 | 0.659 | 0.583 | 0.563 | |||
| STOP-Bang ≥5 previous(p) vs. new(n) cut-off points | Sensitivityp | 72.5 | 77.2 | 81.5 | 70.3 | 74.9 | 80.7 | 59.2 | 70.8 | 77.9 | ||
| Specificityp | 71.7 | 60.1 | 47 | 63.1 | 52.2 | 45.4 | 84.8 | 70.5 | 62.2 | |||
| AUCp | 0.591 | 0.687 | 0.643 | 0.667 | 0.636 | 0.630 | 0.720 | 0.707 | 0.701 | |||
| Sensitivityn | 79.3 | 83.8 | 87.6 | 73.0 | 77.2 | 82.9 | 73.0 | 77.2 | 82.9 | |||
| Specificityn | 66.7 | 53.8 | 40.0 | 60.2 | 48.2 | 42.1 | 60.2 | 48.2 | 42.1 | |||
| AUCn | 0.730 | 0.688 | 0.638 | 0.666 | 0.627 | 0.625 | 0.759 | 0.691 | 0.677 | |||
Abbreviations: AUC, area under the curve; AHI, apnea-hypopnea index.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
