Association between weight-adjusted waist index and obstructive sleep apnea among adults in the United States: data from NHANES (2005–2008 and 2015–2018)

Abstract

Objective

We aimed to examine the relationship between the weight-adjusted waist index (WWI) and obstructive sleep apnea (OSA), a condition often caused by obesity, which remains unclear.

Methods

In this cross-sectional study, we analyzed data from the National Health and Nutrition Examination Survey among adults in the United States (US) aged 20 to 65 years, covering the periods 2005 to 2008 and 2015 to 2018. The study included 8278 participants; we used multivariate logistic regression, restricted cubic splines, and subgroup analyses to explore the relationship between WWI and OSA.

Results

After adjusting for all covariates, each unit increase in WWI was associated with a 30% increase in OSA prevalence (odds ratio = 1.30, 95% confidence interval: 1.20–1.40).

Conclusions

These findings suggest that WWI, an index reflecting abdominal obesity, can provide important insights into OSA risk assessment. Its strong association with OSA highlights its potential utility in predicting OSA prevalence, particularly among diverse subpopulations. The WWI was associated with a higher prevalence of OSA among US adults and may serve as a valuable tool for risk assessment, early screening, and intervention strategies in clinical practice.

Introduction

Understanding sleep begins with recognizing its six primary functions: reducing caloric expenditure, enhancing immune function, restoring neuronal connectivity, repairing brain performance, promoting cerebral lymphatic circulation, and maintaining synaptic function. ¹ Sleep deprivation can manifest in various ways, including overweight, cognitive dysfunction, and immune system disorders. ² In modern society, where development is rapid, the human brain processes vast amounts of information daily; thus, adequate and healthy sleep is crucial for maintaining overall health. ³ The National Sleep Foundation recommends that individuals aged 18 to 64 years get 7 to 9 hours of sleep daily to maintain emotional stability, physical health, and cognitive clarity. ⁴ Findings from the Centers for Disease Control and Prevention (CDC) indicate that the number of adults sleeping fewer than 6 hours a day increased significantly between 1985 and 2012, rising from 38.6 million to 70.1 million. The CDC considers sleep deprivation a public health epidemic. ⁵

In 2019, nearly 1 billion adults aged 30 to 69 years worldwide had mild to severe obstructive sleep apnea (OSA), with 425 million of these individuals representing moderate to severe cases. ⁶ In recent years, attention has gradually shifted from issues of hygiene and malnutrition to chronic diseases, ⁷ such as obesity, population aging, and OSA. ⁸ It is recognized that OSA poses a significant threat to global health. ⁹ OSA is commonly characterized by frequent episodes of apnea and hypopnea during sleep, resulting in sleep fragmentation and chronic intermittent hypoxia. These conditions can lead to serious health issues, including diabetes, hypertension, sudden death, and coronary heart disease in patients. ¹⁰

Obesity is the most critical risk factor for the development of OSA, with the incidence of OSA in obese patients increasing each year. ¹¹ The most common condition among patients with OSA is abdominal obesity, which significantly increases the risk of developing OSA. ¹² The weight-adjusted waist index (WWI), which standardizes waist circumference by weight, offers a more accurate evaluation of abdominal fat distribution, a key factor closely linked to OSA risk. Furthermore, the WWI demonstrates greater accuracy in identifying risks associated with metabolic syndrome and cardiovascular diseases, emphasizing its importance as a tool for assessing obesity-related health outcomes. ¹³ However, its association with OSA remains unclear. We aimed to explore the relationship between the WWI and OSA, providing guidance for clinicians to more effectively prevent and treat OSA.

Methods

Study population

In this study, we used publicly available data from the National Health and Nutrition Examination Survey (NHANES). NHANES is a nationally representative survey among adults in the United States (US) approved by the Research Ethics Review Board of the National Center for Health Statistics (NCHS). The research protocol for the 2005 to 2006 and 2007 to 2008 NHANES data is Protocol #2005-06. The research protocol for the 2015 to 2016 and 2017 to 2018 NHANES data is Protocol #2011-17. This study was conducted in accordance with the Helsinki Declaration of 1975, as revised in 2013. All participants provided written informed consent before participating in the data collection process, which included demographics, diet, physical examinations, laboratory tests, and questionnaires conducted by the Centers for Disease Control and Prevention (CDC). All patient data used in this study were de-identified in accordance with NHANES protocols. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. ¹⁴ Because the 2009 to 2014 cycle did not include OSA data, we used datasets from the 2005 to 2008 and 2015 to 2018 cycles for this study. We only included participants aged 20 to 65 years who had both WWI and OSA data available (Figure 1).

Figure 1.

Flow chart of participant selection. NHANES, National Health and Nutrition Examination Survey.

WWI

In this study, the independent variable WWI is a continuous variable calculated based on the formula: waist circumference divided by the square root of body weight. ¹⁵ Waist circumference is measured in centimeters and body weight in kilograms.

OSA

The dependent variable OSA in this study is defined as the response to any one of three binary questions: (1) feeling excessively sleepy during the day at least 16 to 30 times per month while sleeping approximately 7 hours or more per night on weekdays or workdays; (2) reporting apnea, snoring, or gasping for air more than three nights per week; (3) snoring at least three nights per week. ¹⁶

Covariates

Covariates included age, sex, race and ethnicity, family income to poverty ratio (PIR), education level, diabetes status, hypertension, smoking, alcohol use, body mass index (BMI), and engaging in moderate physical activity. Participants were classified as having hypertension with systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥80 mmHg, based on the average of three resting measurements. Participants who self-reported a prior diagnosis of hypertension were also included in this classification. ¹⁷ Diabetes was diagnosed based on any of the following criteria: fasting plasma glucose ≥7.0 mmol/L, glycated hemoglobin ≥6.5%, random plasma glucose ≥11.1 mmol/L, current use of antidiabetic medications or insulin, or self-reported physician-diagnosed diabetes. ¹⁸ Smoking status was determined through self-report and classified into three categories: never smoker was defined as an individual who had smoked fewer than 100 cigarettes in their lifetime; former smoker was defined as one who had smoked more than 100 cigarettes but had quit; and current smoker was defined as one who smoked daily or occasionally at the time of the survey. ¹⁹ Alcohol consumption was evaluated using self-reported data from the NHANES Computer-Assisted Personal Interview questionnaire, which recorded the frequency of alcohol use over the past year. Based on their monthly drinking frequency, participants were classified into four groups. ²⁰ Moderate physical activity was evaluated using the NHANES physical activity questionnaire. Activities classified as moderate intensity included those with a metabolic equivalent value of 4. Participants reporting at least 150 minutes of moderate-intensity activity per week were categorized as engaging in moderate physical activity. ²¹ Detailed measurement information on confounding factors can be accessed through the official website at http://www.cdc.gov/nchs/nhanes/.

Statistical analyses

This was a cross-sectional study. We used the R statistical software package (http://www.r-project.org) for data analysis, with a significance threshold set at p < 0.05. Sampling weights were calculated strictly following NCHS design guidelines. Multivariable logistic regression was used to analyze the linear relationship between the WWI and OSA rates. Restricted cubic splines were used to investigate the nonlinear association between WWI and OSA. Three multivariable regression models were analyzed. Model 1 did not adjust for covariates. Model 2 adjusted for race and ethnicity, sex, and age. Model 3 further adjusted for education level, family PIR, diabetes, smoking, alcohol, hypertension, physical activity, and BMI, in addition to the adjustments made in Model 2. Calibration plots were used to assess the agreement between predicted probabilities and observed outcomes, applying bootstrap resampling methods for validation.

Results

Baseline characteristics

A total of 8278 participants were included in this study, with an average age of 41.76 ± 12.81 years. Of these, 4210 were men (50.86%) and 4068 were women (49.14%). Among all participants, 61.89% were non-Hispanic, 17.89% were Mexican-American, and 20.22% were individuals of other racial and ethnic backgrounds. Among the 8278 participants, 4303 had OSA, with a prevalence of 51.98%. Compared with the lowest quartile of WWI, those in the highest quartile were more likely to be women and aged 40 to 65 years; more likely to have less than a high school education, a non-Hispanic racial and ethnic background, obesity, a moderate household income, diabetes, and hypertension; and more likely to consume 1 to 5 alcoholic drinks per month. Conversely, participants in the highest quartile exhibited lower levels of physical activity and smoking (WWI quartile Q1: <10.31; WWI quartile Q2: 10.31–10.86; WWI quartile Q3: 10.86–11.42; WWI quartile Q4: >11.42; Table 1).

Table 1.

Demographic characteristics of participants according to the WWI among adults in the United States aged 20–65 years.

Characteristics	WWI				p-value
	Q1 (<10.31) N = 2122	Q2 (10.31–10.86) N = 2040	Q3 (10.86–11.42) N = 2070	Q4 (>11.42) N = 2046
Age (mean ± SD) (years)	34.95 ± 11.64	41.76 ± 11.72	44.71 ± 12.16	47.64 ± 12.23	<0.001
Age (years)					<0.001
20–40	69.45	47.49	37.76	27.9
40–65	30.55	52.51	62.24	72.1
Sex (%)					<0.001
Male	67.44	63.68	50.10	21.65
Female	32.56	36.32	49.90	78.35
Race and ethnicity (%)					<0.001
Non-Hispanic	76.05	61.90	55.85	53.55
Mexican American	7.71	16.49	22.29	25.26
Other races and ethnicities	16.24	21.61	21.86	21.19
Education level (%)					<0.001
Less than high school	86.19	85.84	83.19	79.30
High school and above	5.55	6.01	5.85	8.07
Moderate physical activity (%)					<0.001
Yes	51.04	54.31	52.66	48.48
No	48.96	45.69	47.34	51.52
Family PIR (%)					<0.001
Low-income	17.54	17.73	18.78	25.95
Moderate-income	48.77	50.71	53.81	51
High-income	33.69	31.56	27.41	23.05
BMI (kg/m2) (%)					<0.001
Normal weight (<25 kg/m2)	60.43	31.38	18.88	8.27
Overweight (25–30 kg/m2)	30.02	39.76	36.49	22.32
Obesity (>30 kg/m2)	9.55	28.86	44.63	69.41
Diabetes (%)					<0.001
Yes	14.14	8.53	6.38	5.23
No	85.86	91.47	93.62	94.77
Smoking status (%)					<0.001
Current smoker	22.1	24.41	25.65	23.02
Former smoker	22.95	22.3	20.05	13.49
Nonsmoker	54.95	53.28	54.3	63.49
Alcohol intake (%)					<0.001
Non-drinker	20.14	24.74	28.34	35.68
1–5 drinks/month	50.02	50.43	50.34	50.37
5–10 drinks/month	12.63	9.47	8.44	5.63
>10 drinks/month	17.20	15.36	12.88	8.32
Hypertension (%)					<0.001
Yes	40.25	28.24	23.86	17.5
No	59.75	71.76	76.14	82.5
OSA (%)					<0.001
Yes	36.14	51.37	55.61	65.05
No	63.86	48.63	44.39	34.95

Mean ± standard deviation (SD) for continuous variables: Kruskal–Wallis rank-sum test for complex survey samples; percent for categorical variables. The p-values were calculated using the chi-squared test with Rao–Scott second-order correction.

PIR, poverty income ratio; BMI, body mass index; Q, quartile; OSA, obstructive sleep apnea; WWI, weight-adjusted-waist index.

Association between WWI and OSA

Table 2 presents the results of weighted multivariable logistic regression analysis for the crude model (Model 1), partially adjusted model (Model 2), and fully adjusted model (Model 3). WWI and OSA showed a positive correlation across all three models, with the odds ratio (ORs) and (95% confidence interval [CI]) for Model 1: 1.70 (1.60–1.80), p < 0.001; Model 2: 1.75 (1.65–1.87), p < 0.001; and Model 3: 1.30 (1.20–1.40), p < 0.001. In Model 3, for every unit increase in WWI, the risk of OSA was increased by 30%. The positive correlation between the four quartiles of WWI and OSA was statistically significant in all three models, with the p for trend all <0.001. Additionally, we analyzed the smooth curve fitted by the restricted cubic spline plot, which confirmed the positive linear association between WWI and OSA (Figure 2).

Table 2.

Associations between WWI and OSA.

	Model 1OR (95% CI)	Model 2OR (95% CI)	Model 3OR (95% CI)
	p-value	p-value	p-value
WWI	1.70 (1.60–1.80)	1.75 (1.65–1.87)	1.30 (1.20–1.40)
Continuous	<0.001	<0.001	<0.001
Categories
Quartile 1	reference	reference	reference
Quartile 2	1.87 (1.65–2.11)	1.81 (1.59–2.06)	1.38 (1.21–1.58)
	<0.001	<0.001	<0.001
Quartile 3	2.21 (1.95–2.51)	2.14 (1.87–2.44)	1.38 (1.19–1.60c)
	<0.001	<0.001	<0.001
Quartile 4	3.29 (2.89–3.74)	3.44 (2.99–3.97)	1.81 (1.53–2.13)
	<0.001	<0.001	<0.001
p for trend	<0.001	<0.001	<0.001

Model 1: no covariates were adjusted.

Model 2: age, sex, race and ethnicity were adjusted.

Model 3: age, sex, race and ethnicity, education level, PIR, BMI, moderate activity, smoke, alcohol, diabetes, and hypertension were adjusted.

OSA, obstructive sleep apnea; WWI, weight-adjusted-waist index; OR, odds ratio; CI, confidence interval.

Figure 2.

Linear relationship between the weight-adjusted waist index (WWI) and obstructive sleep apnea (OSA) (solid red line) fitted by the smoothed curve.

Subgroup analyses

Figure 3 presents the results of subgroup analysis and interaction tests, which were stratified by age, sex, race and ethnicity, BMI, smoking status, moderate physical activity, diabetes, hypertension, alcohol intake, education level, and family PIR. In this analysis, we aimed to examine the stability and population differences in the association between the WWI and OSA across different subgroups.

Figure 3.

Results of subgroup analysis for the association between the weight-adjusted waist index (WWI) and obstructive sleep apnea (OSA). OR, odds ratio; CI, confidence interval; BMI, body mass index.

The positive correlation between WWI and OSA remained consistent across most subgroups. However, in the subgroups defined by diabetes, hypertension, and alcohol intake (>10 drinks/month), this positive correlation was not stable. This suggests that the association between WWI and OSA may not be applicable to individuals with diabetes, hypertension, or high alcohol consumption. These findings highlight the need to consider these specific conditions when evaluating the relationship between the WWI and OSA. The calibration plot indicated strong alignment between the predicted probabilities of the WWI-based model and the observed probabilities across most ranges (Figure 4).

Figure 4.

Calibration curve of Model 3.

Discussion

In the present cross-sectional study among 8278 adults in the US aged 20 to 65 years, we found a positive correlation between the WWI and incidence of OSA. Participants with a higher WWI had a greater incidence of OSA. Even when WWI was categorized into quartiles (Q1–Q4), the positive correlation with OSA remained consistent. Subgroup analyses and interaction tests confirmed that the positive correlation between WWI and OSA remained stable across all stratified subgroups except those with diabetes, hypertension, and consumption of >10 alcoholic drinks/month. The calibration plot demonstrated the reliability of the WWI-based model in predicting OSA risk, particularly among low-risk groups, highlighting its potential as a clinically valuable tool. This study provides the first evidence that elevated WWI may be an independent risk factor for OSA, highlighting the importance of WWI in the diagnosis and treatment of OSA.

Obesity is a leading cause of OSA, a common organic sleep disorder. ²² Research shows that 70% of patients with OSA are clinically diagnosed as obese. ²³ Furthermore, 60% of obese individuals have OSA, indicating a strong correlation between the two conditions. ²⁴ In fact, two to three of every three people with OSA are obese. ²⁵ Morbidly obese patients exhibit a higher prevalence and severity of OSA. ²⁶ Studies have shown that for every 10% increase in body weight, the prevalence of moderate or severe OSA increases sixfold. ²⁷ For each standard deviation increase in BMI, the incidence of OSA increases fourfold. ²⁸ With BMI >29 kg/m², the risk of developing OSA increases tenfold. ²⁹

There are three main reasons why obesity induces OSA; the first is owing to the direct impact of obesity. The volume of soft tissue in the airway wall is increased in obese individuals, compressing the pharyngeal airway and causing it to narrow. ³⁰ Whereas the increased activity of upper airway muscles in obese individuals during wakefulness can compensate for the narrowing caused by soft tissue accumulation, this compensatory mechanism fails during sleep. Consequently, the upper airway narrows, and the accumulated soft tissue obstructs it, resulting in OSA. ³¹ The second reason is owing to the indirect impact of obesity on the upper airway. Increased abdominal fat during supine sleep compresses lung capacity. The reduction in lung capacity leads to decreased muscle tone of the pharyngeal sidewalls and lower tracheal/longitudinal traction, causing the upper airway to collapse. This further exacerbates airway narrowing during sleep. ³² Finally, obesity impacts the neuro-mechanical function of the upper airway. The collapse of the pharyngeal airway stimulates the high expression of inflammatory genes, leading to the overexpression of macrophage inflammatory protein-2, tumor necrosis factor-α, interleukin-1β, and P-selectin. This triggers an inflammatory cascade in the mucosal tissues of the upper airway, causing excessive immune cell infiltration and post-inflammatory remodeling of the extracellular matrix in the mucosal tissues. This results in inflammatory changes in the ultrastructure of the pharyngeal upper airway and sensory deficits in the nervous system. These sensory deficits weaken the protective effects of the negative pressure reflex within the pharyngeal cavity. ³³ Leptin, a satiety hormone secreted by adipose tissue, often shows resistance in obese patients, leading to impaired neuro-mechanical control of the respiratory center. As a result, the neuromuscular control system loses its ability to compensate for upper airway collapse, and the decreased respiratory drive increases the susceptibility to OSA. ³⁴

The WWI is an indicator of abdominal obesity. It is calculated as the waist-to-weight ratio, adjusted for body weight, to help distinguish between muscle mass and fat mass. ³⁵ Compared with BMI, the WWI is more closely associated with visceral fat accumulation and can exclude the influence of overall body weight, making it a more accurate predictor of central obesity. ³⁶ The WWI affects OSA in three ways. First, the WWI is positively correlated with an increase in upper airway neck circumference and fat tissue volume. The accumulation of neck fat inhibits pharyngeal nerve reflexes, obstructing the mechanical control function of the pharyngeal nerves, which mediates the collapse of the pharyngeal airway, passively increasing the susceptibility to OSA syndrome. Second, individuals with a higher WWI and central obesity accumulate excessive fat in the abdomen and chest, reducing lung compliance. This leads to a decrease in expiratory reserve volume and a reduction in resting lung volume, resulting in insufficient tracheal/longitudinal traction force. Finally, individuals with a higher WWI not only lack a passive respiratory drive but also have unstable feedback loop functions of the respiratory control system’s “gain devices,” reducing lung ventilation. ³⁷ This physical mechanical effect induces OSA.

The incidence of obesity and related diseases is increasing annually in modern societies, making it crucial to accurately assess the prevalence of OSA and identify individuals at risk. This study is the first to analyze the association between WWI and the incidence of OSA, providing clinically applicable results for assessing OSA risk. However, this study has several limitations. Our cross-sectional study used data provided by NHANES and we selected covariates that could potentially affect the results. There are many unconsidered variables that might also affect the results, which were not included in this study owing to space constraints. Furthermore, the inherent limitations of cross-sectional studies prevented the complete elucidation of the causal relationship between the WWI and OSA.

Conclusion

The results of this cross-sectional study revealed a positive correlation between the WWI and the incidence of OSA among US adults aged 20 to 65 years. These findings imply that managing WWI within a healthy range may help reduce the prevalence of OSA. To confirm these results, further case-control studies are needed.

Data availability

The survey data are publicly available on the internet for data users and researchers worldwide (www.cdc.gov/nchs/nhanes/).