Rethinking neck circumference in STOP-BANG for Asian OSA

Abstract

Introduction:

The STOP-BANG questionnaire is a simple screening tool for obstructive sleep apnoea (OSA) which has been validated for use in Asian populations. The original cut-off values for neck circumference were defined based on Caucasian populations. This study aims to validate the neck circumference component of the STOP-BANG score in Asians.

Methods:

A retrospective analysis of all patients suspected of having OSA who underwent a diagnostic polysomnography at a tertiary sleep centre in Singapore over the course of 1 year was conducted. Neck circumference was measured and correlated to the presence of OSA. A receiver operating curve was performed to obtain the optimal cut-off value for screening OSA in Asian patients.

Results:

591 Asian subjects with OSA were analysed. The optimal neck circumference predictive for OSA was found to be 39 cm in males and 35 cm in females, with a sensitivity of 83.4% and 84.6% respectively. The existing STOP-BANG cut-off of 40 cm had 71% and 32% sensitivity in Asian males and females, respectively.

Conclusion:

The STOP-BANG cut-off value for neck circumference for screening for sleep apnoea in Asian males can be maintained at 40 cm, but the neck circumference value used for screening Asian females should be lowered to 35 cm.

Keywords

Obstructive sleep apnoea sleep disordered breathing neck circumference screening STOP-BANG questionnaire

Introduction

Obstructive sleep apnoea (OSA) is a common disorder characterised by repetitive partial and complete upper airway obstructions during sleep resulting in intermittent hypoxaemia, sympathetic nervous system output spikes and sleep arousals.¹ As compared to normal individuals, epidemiological studies have shown that patients with OSA have increased cardiovascular and cerebrovascular morbidity, excessive daytime sleepiness, poor neurocognitive function and work performance, increased motor vehicular accidents, and reduced quality of life.^2–5

OSA is readily treatable in patients, but identifying patients remains a challenge. The gold standard for diagnosis of OSA is overnight polysomnography (PSG), which is costly and time consuming to perform as a screening procedure. Hence, numerous screening questionnaires such as the STOP-BANG questionnaire, Berlin questionnaire, American Society of Anesthesiology Checklist and Wisconsin questionnaire have been developed for PSG eligibility.

The STOP-BANG questionnaire was first described by Chung et al. in 2008.⁶ It consists of four questions regarding snoring, tiredness during daytime, observed apnoea and high blood pressure, and four objective measurements – body mass index (BMI) greater than 35 kg/m², age above 50 years, neck circumference greater than 40 cm and male gender.

The questionnaire has been validated in Asia.^7–9 A lower neck circumference and BMI cut-off of 36 cm and 25 kg/m², respectively, were suggested for the Thai population.⁸ However, lower BMI cut-offs did not affect the overall STOP-BANG performance in the Singapore population. Neck circumference cut-off was not evaluated in the previous Singapore study. This paper aims to validate the STOP-BANG cut-off for neck circumference in the Asian population.

Methods

All cases that were referred to a tertiary sleep centre in Singapore for diagnostic PSG over a year were analysed retrospectively. Ethics approval was obtained from the institution’s ethical review board. All full night studies and diagnostic component of split night studies were included. Non-diagnostic studies performed for bi-level positive airway pressure (BIPAP) or continuous positive airway pressure (CPAP) titration and post-intervention studies for treatment such as dental splints or corrective sleep surgery were excluded from analysis. Age, gender, neck circumference, height, weight, BMI and apnoea-hypopnoea index (AHI) were recorded and analysed. Apnoea and hypopnoea were scored based on the 2007 American Academy of Sleep Medicine manual.¹⁰ Neck circumference was measured at the level of the cricothyroid membrane by a sleep technician on the evening of the sleep study. Nasal and oral airflow, electrocardiogram, chest wall and abdominal movements, oxygen saturation, snoring were recorded in every sleep study.

OSA was defined based on AHI, with 5 to less than 15 for mild, 15 to less than 30 for moderate and 30 or more for severe. Patients with an AHI < 5 were considered to have no OSA.

Data were analysed using Statistical Package for the Social Sciences (SPSS, Version 15, IBM Corporation, 2017). Spearman’s rank correlation was performed to assess the relationship between neck circumference and AHI. Multi-variate analysis of variance (MANOVA) was carried out to determine interactions between variables. Sensitivity and specificity of cut-off values which predict the presence of OSA were evaluated by a receiver operating characteristic (ROC) analysis. The Youden index (sensitivity + specificity − 1) was used to obtain the optimal cut-off value with the maximal sensitivity and specificity.

Results

A total of 1003 polysomnograms were performed at our sleep centre over the 1 year study period. 591 met our inclusion criteria. 298 post treatment follow-up studies and 98 titration studies were excluded.

The demographics and characteristics of the study subjects are described in Tables 1 and 2. Pearson Chi-squared test showed a statistically significant difference (p < 0.001) between mean AHI in males and females. Therefore, subgroup analysis was performed by gender.

Table 1.

Characteristics of subjects.

	Total (n = 591)	Male (n = 435)	Female (n = 156)
Mean age ± SD	46.3 ± 13.5	45.2 ±13.5	49.2 ±13.1
Mean height ± SD (cm)	166.2 ± 8.7	169.6 ± 6.7	156.9 ± 6.6
Mean weight ± SD (kg)	77.4 ± 18.2	80.3 ± 17.8	69.1 ±16.6
Mean BMI ± SD (kg/m²)	28.0 ± 6.1	27.9 ± 5.9	28.1 ± 6.8
Mean neck circumference ± SD (cm)	39.9 ± 4.4	41.1 ± 4.0	36.6 ± 3.6
Mean AHI ± SD (events/h)	25.4 ± 29.1	28.0 ± 29.5	18.2 ± 27.1
Race
Chinese	487	359	128
Malay	49	34	15
Indian	45	33	12
Other	10	9	1

SD – one standard deviation.

Table 2.

Average neck circumference with standard deviation.

	No OSA (AHI < 5)	OSA(AHI > 5)	Mild OSA (AHI 5–15)	Moderate OSA (AHI 15–30)	Severe OSA (AHI > 30)
All	37.1 ± 3.3	41.0 ± 4.3	39.6 ± 3.5	40.7 ± 3.9	42.3 ± 4.6
Male	38.7 ± 2.6	41.0 ± 4.3	40.5 ± 3.2	41.3 ± 3.6	42.9 ± 4.59
Female	34.6 ± 2.7	38.0 ± 3.5	37.1 ± 3.1	37.5 ± 3.6	39.4 ± 3.5

Normal studies (AHI < 5.0/h) were found in 27.6% (n = 163) subjects, 26.2% (n = 155) had mild OSA, 15.4% (n = 91) had moderate OSA and 30.8% (n = 182) had severe OSA. Subjects with normal studies had a mean neck circumference of 37.1 cm, whereas subjects with mild, moderate and severe OSA had a mean neck circumference of 39.6 cm, 40.7 cm and 42.3 cm, respectively.

Spearman’s rank correlation analysis confirmed a positive relationship between neck circumference and AHI (r = 0.507). An ROC analysis of neck circumference against subjects with OSA (AHI > 5) showed an area under the curve (AUC) of 0.768, 0.749 and 0.776 for all subjects, males and females, respectively (Figures 1 –3).

Figure 1.

Receiver operating characteristic curve for neck circumference against AHI > 5 (area under curve = 0.768).

Figure 2.

Receiver operating characteristic curve for neck circumference against AHI > 5 in male subjects (area under curve = 0.749).

Figure 3.

Receiver operating characteristic curve for neck circumference against AHI > 5 in female subjects (area under curve = 0.776).

The cut-off values yielding the optimal sensitivity and specificity based on the Youden index was found to be 39 cm in males (83.4% sensitivity, 53.1% specificity) and 35 cm in females (84.6% sensitivity, 56.9% specificity). Sensitivity and specificity analysis was repeated for other common cut-off values of 40 cm, 41 cm and 43 cm (Table 3).¹¹

Table 3.

Predictive parameters of neck circumference values for presence of OSA.

Cut-off neck circumference (cm)	Sensitivity (%)	Specificity (%)	Positive predictive value (%)	Negative predictive value (%)
All
34	49.2	17.2	75.5	6.1
39	74.0	68.1	85.9	50.0
40	62.9	75.4	87.1	43.6
41	53.3	84.0	89.7	40.7
43	32.5	92.6	92.1	34.3
Males
39	83.4	53.1	85.9	48.1
40	71.2	63.3	87.0	39.0
41	61.1	75.5	89.6	36.0
43	38.3	88.8	92.1	29.5
Females
34	90.1	43	68.9	75.7
35	84.6	56.9	73.3	72.5
39	39.6	90.8	85.7	51.8
40	31.9	93.8	87.9	49.6
41	24.2	96.9	91.6	47.7
43	11.1	98.5	90.9	44.1

Kruskal-Wallis test of variance showed no statistical difference in mean AHI between races (p = 0.067). MANOVA showed no significant interaction between age or height with weight or neck circumference.

BMI (r = 0.498) showed a positive correlation with AHI on Spearman’s rank correlation analysis. Correcting for BMI by using neck circumference to BMI ratio rather than neck circumference alone did not improve the correlation (r = 0.281), and did not improve the ROC (AUC = 0.356).

Discussion

The relationship between neck circumference and OSA was first described by Davies and colleagues in 1990.^12,13 A neck circumference cut-off of 16.75 inches/42.5 cm in a 1.78 m tall male was found to be 87% sensitive and 79% specific for significant OSA. This figure rounded up to 17 inches/43 cm in males has been adopted in the American Academy of Sleep Medicine guidelines as the neck circumference cut-off suggestive of OSA.¹¹

The Caucasian cut-off value for neck circumference of 43 cm in males and 41 cm in females has poor sensitivity of 38.3% and 24.1%, respectively, for screening for OSA in Asians. This finding is not surprising considering that structural differences have been described between Asian and Caucasian airways, and Asians have a narrower neck circumference and lower BMI compared to Caucasians for the same severity of OSA.¹⁴

In the first STOP publication,⁶ subjects with AHI < 5 had an average neck circumference of 36 cm, while subjects with AHI > 5 had a neck circumference of 40 cm. This was similar to our population demographic with an average neck circumference of 37 cm in patients without OSA, and 41 cm in patients with OSA. The racial distribution of the subjects in the STOP publication was not described. The city of Toronto, where the STOP study was based, is known to have a 40% proportion of ethnic Asians in the population,¹⁵ which could account for a value similar to ours.

A cut-off value of 39 cm gave the best sensitivity and specificity of 83.4% and 53.1%, respectively, in males. Using the STOP-BANG cut-off of 40 cm would decrease the sensitivity to 71%. However, the existing cut-off is simple to remember, convenient to use and already validated when used in conjunction with other STOP-BANG parameters in the Asian context.⁷ For these reasons, we propose to maintain the current cut-off values for males at the cost of a slightly lower sensitivity.

Applying the 40 cm cut-off value for females would yield a low sensitivity of 31.9%. The optimal cut-off for the females within this cohort was 35 cm, which is closer to the proposed Thai cut-off of 36 cm.⁸ Females are known to have smaller neck circumference,¹⁶ lesser soft tissue bulk and different distribution of fat in the neck compared to males.¹⁷ Thus the cut-off neck circumference value used for screening females should be lowered.

Limitations

There were some inherent limitations to this study. The cohort was recruited from patients that were referred for sleep study by a physician. These patients were already suspected to have a sleep disorder, thus the data may not be reflective of population norms. Historical cut-off norms were also based on sleep study cohorts,¹³ rather than the general population. Further validation with a large-scale population based study would be ideal, but would be logistically challenging due to the cost and limited availability of sleep studies.

This study was based on retrospective data. Current definitions for apnoea and hypopnoea have changed with updated scoring guidelines.

There were only 156 females in the study, which was far lower than the number of males. Further study with a larger cohort and focus on OSA screening for females is warranted to further refine available screening tools.

Conclusion

Footnotes

Authors’ contributions

The first author researched literature, conceived the study, and analysed data. Both authors were involved in gaining ethical approval, editing and approving the final version of the manuscript.

Availability of data and materials

The datasets generated and/or analysed during the current study is held by the corresponding author. Patient data has not been made available publicly to protect patient confidentiality.

Conflict of interest

The authors declare no conflict of interests.

Ethical approval

Ethical approval was obtained from the Singhealth Centralized Institutional Review Board (approval number CIRB 2012/334/D).

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Informed consent

Informed consent was not sought for the present study because analysis made on anonymized data posed no more than minimal risk to the participants. Permission for waiver of informed consent was granted by the Singhealth Centralized Institutional Review Board (approval number CIRB 2012/334/D).

ORCID iD

John Ming-Ren Loh

References

Guilleminault

Tilkian

Dement

WC.

The sleep apnea syndromes. Annu Rev Med 1976; 27: 465–484.

Engleman

Douglas

NJ.

Sleep∙ 4: sleepiness, cognitive function, and quality of life in obstructive apnoea/hypopnoea syndrome. Thorax 2004; 59: 618–622.

Sassani

Findley

Kryger

et al . Reducing motor-vehicle collisions, costs, and fatalities by treating obstructive sleep apnea syndrome. Sleep 2004; 27: 453–458.

Moyer

Sonnad

Garetz

et al . Quality of life in obstructive sleep apnea: a systematic review of the literature. Sleep Med 2001; 2: 477–491.

Hack

Choi

Vijayapalan

et al . Comparison of the effects of sleep deprivation, alcohol and obstructive sleep apnoea (OSA) on simulated steering performance. Respir Med 2001; 95: 594–601.

Chung

Yegneswaran

Liao

et al . STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology 2008; 108: 812–821.

Ong

Raudha

Fook-Chong

et al . Simplifying STOP-BANG: use of a simple questionnaire to screen for OSA in an Asian population. Sleep Breath 2010; 14: 371–376.

Pavarangkul

Jungtrakul

Chaobangprom

et al . The STOP-BANG questionnaire as a screening tool for obstructive sleep apnea-induced hypertension in Asian population. Neurol Int 2016; 8: 9–11.

Tan

Yin

JDC

Tan

LWL

et al . Predicting obstructive sleep apnea using the STOP-Bang questionnaire in the general population. Sleep Med 2016; 27–28: 66–71.

10.

Iber

Ancoli-Israel

Chesson

et al . The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications. 1st ed. Westchester, IL: American Academy of Sleep Medicine, 2007.

11.

Epstein

Kristo

Strollo

et al . Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med 2009; 5: 263–276.

12.

Davies

Stradling

JR.

The relationship between neck circumference, radiographic pharyngeal anatomy, and the obstructive sleep apnoea syndrome. Eur Respir J 1990; 3: 509–514.

13.

Davies

Ali

Stradling

JR.

Neck circumference and other clinical features in the diagnosis of the obstructive sleep apnoea syndrome. Thorax 1992; 47: 101–105.

14.

Lee

RWW

Vasudavan

Hui

et al . Differences in craniofacial structures and obesity in Caucasian and Chinese patients with obstructive sleep apnea. Sleep 2010; 33: 1075–1080.

15.

Statistics Canada. Immigration and ethnocultural diversity in Canada, www12.statcan.gc.ca/nhs-enm/2011/as-sa/99–010-x/99–010-x2011001-eng.cfm (2018, accessed 16 June 2018).

16.

Dancey

Hanly

Soong

et al . Gender differences in sleep apnea: the role of neck circumference. Chest 2003; 123: 1544–1550.

17.

Whittle

Marshall

Mortimore

et al . Neck soft tissue and fat distribution: comparison between normal men and women by magnetic resonance imaging. Thorax 1999; 54: 323–328.