Association Between Obstructive Sleep Apnea and Regional Fat: A Cross-Sectional Analysis of National Health and Nutrition Examination Survey 2015

Abstract

Purpose:

The relationship between regional fat and obstructive sleep apnea (OSA) remains poorly understood. This study seeks to explore the link between regional fat and OSA, utilizing data from the National Health and Nutrition Examination Survey (NHANES).

Methods:

This cross-sectional analysis used NHANES 2015–2018 data. OSA symptoms were assessed through sleep questionnaires. Regional fat mass (FM) was measured using dual-energy X-ray absorptiometry, including trunk, arm, leg, android, gynoid, and abdominal FM. The fat mass index (FMI) was calculated by dividing FM by the square of height. Logistic regression evaluated the association between regional FMI and OSA, with univariate and stratified analyses to identify potential effect modifiers.

Results:

A total of 3,099 participants were included, with 1,595 classified into the OSA group. Significant associations were found between OSA and several regional FMIs, including trunk, arm, leg, android, gynoid, and abdomen. These associations were consistent in males, and in females, leg and gynoid FMI were not linked to OSA. Stratified analyses by race revealed significant associations between OSA and regional FMI indices (trunk, arm, leg, android, gynoid, and abdominal FMI) in non-Hispanic Whites and between OSA and trunk, android, and abdominal FMI in other Hispanics. No associations were observed in the Mexican American or non-Hispanic Black groups. Stratification by body mass index (BMI) indicated distinct profiles: obese individuals (BMI ≥30) showed associations limited to trunk, arm, android, and abdominal FMIs, while nonobese participants (BMI <30) displayed broader associations encompassing all regional FMIs. Both univariate and stratified analyses highlighted abdominal FMI as the strongest predictor of OSA.

Conclusion:

Higher regional FMI, particularly abdominal fat, is associated with an increased risk of OSA, with stronger associations observed in male, White, and nonobese populations.

Introduction

Obstructive sleep apnea (OSA) is a sleep-related breathing disorder characterized by repeated partial or complete obstruction of the pharyngeal airway. Its primary symptoms include snoring, awakenings, breathing pauses, and excessive daytime sleepiness.^1,2 Untreated OSA is associated with severe complications and higher mortality rates,^3,4 posing a significant threat to both individual and global health. A study utilizing advanced statistical modeling techniques reveals that nearly 1 billion adults aged 30–69 worldwide may have OSA, with an estimated 425 million suffering from moderate to severe cases.⁵ In addition to its impact on personal health, the economic cost of OSA is substantial, affecting productivity and public safety both directly and indirectly across individuals, families, and societies.^6
–8

Obesity is the most commonly known risk factor for OSA,¹ and excessive fat deposition may play a mechanistic role in the development and progression of OSA. Fat in the neck region is believed to directly compress the upper airway⁹; accumulation of fat within the thoracic cavity imposes mechanical restrictions on lung expansion, thus reducing lung capacity¹⁰; accumulation of abdominal fat may lead to reduced longitudinal tracheal traction, which increases tracheal pull, thereby narrowing the upper airway.¹¹ However, the specific contributions of different patterns of obesity to the development and severity of OSA remain insufficiently understood.

The traditional anthropometric indices used in OSA research include body mass index (BMI), waist and neck circumferences, neck–waist ratio, waist–hip ratio, and skinfold measurements. These indices have limited accuracy in determining fat distribution.^12,13 Studies have shown that the fat mass index (FMI) is a more accurate reflection of body fat as it is not influenced by other body components, such as lean body mass.^14
–16 Our objective is to investigate the relationship between regional fat distribution and OSA using FMI, with data sourced from the National Health and Nutrition Examination Survey (NHANES) dataset.

Materials and Methods

Data source

This cross-sectional study utilized the publicly accessible NHANES dataset, which does not require further approval from an ethics review board. NHANES is a nationally representative survey conducted by the National Center for Health Statistics in the United States. It employs a stratified multistage probability sampling design to evaluate the relationships between nutrition, health promotion, and disease prevention. The survey, conducted biennially, encompasses demographics, dietary intake, examinations, laboratory tests, and questionnaire components. For more information about the NHANES database, visit http://www.cdc.gov/nhanes.

Study participants

Participants were drawn from the NHANES datasets covering the period from 2015 to 2018. Of the initial cohort of 19,225 individuals, 11,498 were excluded because of the absence of dual-energy X-ray absorptiometry (DXA) data, and 2,542 were omitted because of missing OSA data. Furthermore, 2,086 participants were excluded because of incomplete covariate data. Consequently, the final analysis comprised 3,099 participants, as illustrated in Fig. 1.

FIG. 1.

Participant flow chart.

Obstructive sleep apnea

Based on prior research, a diagnosis of OSA can be established if an individual responds “yes” to at least one of the following three NHANES questions¹⁷: (1) Severe daytime sleepiness despite sleeping at least 7 hours per night, with 16–30 incidents reported; (2) Three or more instances of gasping, grunting, or apneas per week; (3) Snoring occurring three or more times per week.

Fat mass index

Whole-body DXA scans are conducted by trained operators on participants aged 8–59, excluding pregnant women; those who self-report the use of radiographical contrast agents (barium) in the past 7 days; and those who self-report a weight over 450 pounds or a height exceeding 6′5″. The scans use the Hologic Discovery A densitometer (Hologic, Inc., Bedford, Massachusetts, USA). All DXA scans require quality control and analysis to measure soft tissue and bone at various body sites. The current analysis includes fat mass (FM) for the left and right arms, left and right legs, trunk, android, gynoid, and abdominal fat. Arm and leg FM are defined as the sum of the bilateral limb FM. FMI is calculated by dividing the FM (in kilograms) by the square of the height (in meters).

Covariates

In accordance with clinical judgment and previous research,^18,19 the following covariates were included: age (years), gender (male/female), race/ethnicity (non-Hispanic White, non-Hispanic Black, Mexican American, other Hispanic, other), educational attainment (less than high school, high school graduate, some college, or higher), marital status (married/living with a partner, living alone), poverty–income ratio (PIR; calculated as family income divided by the U.S. Department of Health and Human Services poverty threshold specific to household size and survey year), BMI, total cholesterol (TC), and high-density lipoprotein (HDL) levels. Smoking status was categorized into three groups: never (fewer than 100 cigarettes lifetime), former (more than 100 cigarettes lifetime but currently abstinent), and current (more than 100 cigarettes lifetime and currently smoking). Drinking status was classified as drinking four to five or more alcoholic beverages per day. Diabetes and hypertension were self-reported by participants who had been diagnosed by medical professionals.

Statistical analysis

In analyzing NHANES data, this study adhered to the guidelines outlined by NHANES. Interview weights were incorporated into all analyses to ensure nationally representative estimates. To account for standard errors related to the complex survey design, the study utilized primary sampling unit variables and stratification variables.

Descriptive and regression analyses were conducted using weighted samples. Continuous variables, assuming normal distribution, are presented as means ± standard deviations, while categorical variables are reported as counts and weighted percentages. Independent t tests or chi-squared tests were employed to analyze baseline information between the control and OSA groups. Multivariable logistic regression models were used to examine the association between regional FMI and OSA. Three models were constructed: Model 1 was adjusted for age, gender, and race; Model 2 included additional adjustments for education level, PIR, and marital status; Model 3 further accounted for BMI, TC, HDL, smoking status, drinking status, diabetes, and hypertension. Subgroup analyses stratified by gender, race, and BMI were also performed, presented using Model 3. Effect sizes are reported as odds ratios (OR) with corresponding 95% confidence intervals (CI).

All statistical analyses were conducted using R software (version 4.3.2). Statistical significance was determined by a P value <0.05.

Results

Participant weighted baseline characteristics

The study included 3,099 individuals who met the criteria, representing a weighted population of 80,880,482. Baseline characteristics of the participants, categorized by the presence or absence of OSA, are summarized in Table 1. The OSA group, consisting of 1,595 individuals with a mean age of 41 ± 11 years, comprised 41% females and 59% males. This group exhibited significantly higher localized fat indices compared with the non-OSA group. Significant differences were observed between OSA and non-OSA participants in terms of age, gender, race, marital status, educational level, BMI, TC, HDL, smoking status, drinking status, hypertension, and diabetes, with the exception of PIR (P < 0.05).

Table 1.

Weighted Baseline Characteristics

Characteristic	OSA(N = 1,504)	Non-OSA(N = 1,595)	P
Age (years)	37 ± 12	41 ± 11	<0.001
Gender, n (%)			<0.001
Female	784 (52%)	675 (41%)
Male	720 (48%)	920 (59%)
Race, n (%)			0.007
Non-Hispanic White	525 (64%)	525 (60%)
Other/multiracial	329 (11%)	280 (9.7%)
Non-Hispanic Black	278 (9.3%)	300 (10%)
Mexican American	216 (8.8%)	299 (12%)
Other Hispanic	156 (6.7%)	191 (8.1%)
Education status, n (%)			0.004
Below high school	194 (7.0%)	287 (13%)
High school	342 (22%)	360 (23%)
College or above	968 (71%)	948 (64%)
PIR	3.19 (1.62)	2.99 (1.64)	0.058
Marital status, n (%)			<0.001
Married/Living with partner	871 (60%)	1069 (69%)
Living alone	633 (40%)	526 (31%)
BMI			<0.001
≥30	391 (25%)	768 (48%)
<30	1113 (75%)	827 (52%)
TC (md/dL)	187 ± 39	195 ± 42	<0.001
HDL (md/dL)	57 ± 17	51 ± 15	<0.001
Smoking status			<0.001
Current	305 (18%)	415 (25%)
Former	265 (20%)	332 (24%)
Never	934 (62%)	848 (51%)
Drinking status			<0.001
Yes	174 (11%)	291 (19%)
No	1330 (89%)	1304 (81%)
Hypertension
Yes	228 (14%)	455 (28%)
No	1276 (86%)	1140 (72%)
Diabetes			<0.001
Yes	74 (3.2%)	148 (7.0%)
No	1430 (96.8%)	1447 (93.0%)
Trunk FMI	4.05 ± 2.03	5.27 ± 2.37	<0.001
Arms FMI	1.05 ± 0.52	1.29 ± 0.63	<0.001
Legs FMI	3.17 ± 1.44	3.57 ± 1.65	<0.001
Android FMI	0.71 ± 0.41	0.98 ± 0.48	<0.001
Gynoid FMI	1.52 ± 0.65	1.72 ± 0.72	<0.001
Abdominal FMI	0.65 ± 0.35	0.83 ± 0.38	<0.001

BMI, body mass index; FMI, fat mass index; HDL, high-density lipoprotein; OSA, obstructive sleep apnea; PIR, poverty–income ratio; TC, total cholesterol.

Weighted associations between regional FMI and OSA in the entire study population

Three sets of weighted logistic regression models were employed to evaluate the relationships between OSA and regional FMI, as presented in Table 2. All models demonstrated a significant positive association between OSA and FMI across the specified body regions. In the fully adjusted multivariable regression models, the associations were as follows: trunk FMI (OR = 1.26, 95% CI = 1.12–1.42, P = 0.001), arm FMI (OR = 1.95, 95% CI = 1.29–2.97, P = 0.005), leg FMI (OR = 1.21, 95% CI = 1.05–1.40, P = 0.014), android FMI (OR = 3.08, 95% CI = 1.77–5.34, P < 0.001), gynoid FMI (OR = 1.53, 95% CI = 1.08–2.15, P = 0.020), and abdominal FMI (OR = 4.09, 95% CI = 1.90–8.78, P = 0.002). The associations with abdominal FMI were particularly pronounced (Table 2).

Table 2.

Associations Between Regional Fat Mass and OSA in the Entire Study Population

Characteristic	Model 1			Model 2			Model 3
Characteristic	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
Trunk FMI	1.35	1.26–1.44	<0.001	1.35	1.26–1.44	<0.001	1.26	1.12–1.42	0.001
Arms FMI	3.00	2.31–3.90	<0.001	2.95	2.26–3.86	<0.001	1.95	1.29–2.97	0.005
Legs FMI	1.44	1.29–1.61	<0.001	1.43	1.28–1.61	<0.001	1.21	1.05–1.40	0.014
Android FMI	4.24	3.10–5.80	<0.001	4.19	3.03–5.82	<0.001	3.08	1.77–5.34	<0.001
Gynoid FMI	2.30	1.80–2.96	<0.001	2.29	1.77–2.96	<0.001	1.53	1.08–2.15	0.020
Abdominal FMI	6.56	4.25–10.10	<0.001	6.43	4.09–10.10	<0.001	4.09	1.90–8.78	0.002

Model 1 is adjusted for age, gender, and race. Model 2 includes adjustments for age, gender, race, education level, PIR, marital status. Model 3 accounts for age, gender, race, education level, PIR, marital status, BMI, TC, HDL, smoking status, drinking status, diabetes, and hypertension.

CI, confidence interval; OR, odds ratio.

Subgroup analyses

Subgroup analyses were performed to explore potential variations in associations based on gender, race, and BMI.

Gender-stratified analysis

In the gender-stratified analysis, as shown in Table 3, trunk FMI (OR = 1.33, 95% CI = 1.11–1.60, P = 0.005), arm FMI (OR = 3.07, 95% CI = 1.69–5.57, P = 0.001), leg FMI (OR = 1.44, 95% CI = 1.12–1.84, P = 0.008), android FMI (OR = 3.56, 95% CI = 1.59–7.94, P = 0.005), gynoid FMI (OR = 2.13, 95% CI = 1.19–3.81, P = 0.015), and abdominal FMI (OR = 6.45, 95% CI = 2.21–18.90, P = 0.003) were significantly associated with OSA in males. In females, significant associations were observed with trunk FMI (OR = 1.25, 95% CI = 1.07–1.45, P = 0.007), arm FMI (OR = 1.76, 95% CI = 1.05–2.95, P = 0.035), android FMI (OR = 2.97, 95% CI = 1.45–6.08, P = 0.006), and abdominal FMI (OR = 3.52, 95% CI = 1.31–9.49, P = 0.017), and no significant associations were found for leg FMI and gynoid FMI (all P > 0.05).

Table 3.

Associations Between Regional FMI and OSA by Gender

Characteristic	Female			Male
Characteristic	OR	95% CI	P	OR	95% CI	P
Trunk FMI	1.25	1.07–1.45	0.007	1.33	1.11–1.60	0.005
Arm FMI	1.76	1.05–2.95	0.035	3.07	1.69–5.57	0.001
Leg FMI	1.13	0.94–1.35	0.2	1.44	1.12–1.84	0.008
Android FMI	2.97	1.45–6.08	0.006	3.56	1.59–7.94	0.005
Gynoid FMI	1.33	0.87–2.04	0.2	2.13	1.19–3.81	0.015
Abdominal FMI	3.52	1.31–9.49	0.017	6.45	2.21–18.9	0.003

Race-stratified analysis

In the race-stratified analysis, trunk FMI (OR = 1.33, 95% CI = 1.09–1.63, P = 0.009), arm FMI (OR = 2.51, 95% CI = 1.19–5.29, P = 0.019), leg FMI (OR = 1.30, 95% CI = 1.02–1.66, P = 0.039), android FMI (OR = 3.80, 95% CI = 1.54–9.33, P = 0.007), gynoid FMI (OR = 1.81, 95% CI = 1.02–3.21, P = 0.042), and abdominal FMI (OR = 5.46, 95% CI = 1.59–18.80, P = 0.011) were significantly associated with OSA among non-Hispanic Whites. For other Hispanics, significant associations were noted for trunk FMI (OR = 1.36, 95% CI = 1.06–1.74, P = 0.02), android FMI (OR = 3.24, 95% CI = 1.25–8.42, P = 0.02), and abdominal FMI (OR = 4.95, 95% CI = 1.20–20.50, P = 0.031). No significant relationships were observed in the Mexican American and non-Hispanic Black groups (all P > 0.05) (Table 4).

Table 4.

Associations Between Regional FMI and OSA by Race

Characteristic	Non-Hispanic White			Other Hispanic			Mexican American			Non-Hispanic Black			Other/multiracial
Characteristic	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
Trunk FMI	1.33	1.09–1.63	0.009	1.36	1.06–1.74	0.02	1.12	0.93–1.35	0.2	1.03	0.89–1.19	0.7	1.29	1.00–1.67	0.051
Arm FMI	2.51	1.19–5.29	0.019	2.38	0.74–7.68	0.13	1.31	0.75–2.29	0.3	0.93	0.60–1.46	0.7	2.22	0.99–4.95	0.052
Leg FMI	1.30	1.02–1.66	0.039	1.28	0.88–1.88	0.2	1.06	0.82–1.38	0.6	0.93	0.78–1.10	0.3	1.22	0.88–1.70	0.2
Android FMI	3.80	1.54–9.33	0.007	3.24	1.25–8.42	0.02	1.87	0.79–4.44	0.13	1.36	0.62–2.95	0.4	4.46	1.52–13.10	0.01
Gynoid FMI	1.81	1.02–3.21	0.042	1.74	0.81–3.72	0.14	1.08	0.59–1.97	0.8	0.82	0.53–1.28	0.3	1.64	0.82–3.28	0.2
Abdominal FMI	5.46	1.59–18.80	0.011	4.95	1.20–20.50	0.031	2.26	0.66–7.70	0.2	1.18	0.42–3.33	0.7	5.32	1.52–18.60	0.012

BMI-stratified analysis

In individuals with BMI ≥30, trunk FMI (OR = 1.18, 95% CI = 1.05–1.33, P = 0.01), arm FMI (OR = 1.49, 95% CI = 1.06–2.10, P = 0.026), android FMI (OR = 2.13, 95% CI = 1.24–3.66, P = 0.01), and abdominal FMI (OR = 2.60, 95% CI = 1.12–6.05, P = 0.029) were positively associated with OSA, whereas leg/gynoid FMI showed non-significant trends. In contrast, among those with BMI <30, all regional FMIs exhibited stronger significant associations: trunk (OR = 1.37, 95% CI = 1.15–1.63, P = 0.002), arm (OR = 3.11, 95% CI = 1.40–6.94, P = 0.009), leg (OR = 1.32, 95% CI = 1.02–1.71, P = 0.036), android (OR = 5.16, 95% CI = 2.19–12.2, P = 0.001), gynoid (OR = 1.78, 95% CI = 1.02–3.13, P = 0.045), and abdominal FMI (OR = 6.25, 95% CI = 2.25–17.40, P = 0.002), with abdominal fat demonstrating the highest risk elevation in both subgroups (Table 5).

Table 5.

Associations Between Regional FMI and OSA by BMI

Characteristic	BMI ≥ 30			BMI ＜ 30
Characteristic	OR	95% CI	P	OR	95% CI	P
Trunk FMI	1.18	1.05–1.33	0.01	1.37	1.15–1.63	0.002
Arm FMI	1.49	1.06–2.10	0.026	3.11	1.40–6.94	0.009
Leg FMI	1.12	0.98–1.27	0.083	1.32	1.02–1.71	0.036
Android FMI	2.13	1.24–3.66	0.01	5.16	2.19–12.2	0.001
Gynoid FMI	1.28	0.92–1.78	0.13	1.78	1.02–3.13	0.045
Abdominal FMI	2.60	1.12–6.05	0.029	6.25	2.25–17.40	0.002

Discussion

This cross-sectional study analyzed data from a comprehensive, nationally representative community survey conducted in the United States, examining the association between body fat and OSA in a diverse population. To the best of our knowledge, this is the first study to report an association between body fat distribution and OSA within the general U.S. population. We analyzed the regional FMI for the trunk, arm, leg, android, gynoid, and abdominal regions. After controlling for a variety of potential confounding factors, our results indicate that regional FMI in all the aforementioned areas is associated with OSA, with abdominal FMI showing the most significant correlation. This association persists across subgroup analyses by gender, race, and BMI. Stratified analysis further revealed that the relationship between abdominal FMI and OSA is more pronounced in males, the White population, and nonobese population.

Our findings share similarities with those of other studies in this field. In a Mendelian randomization study using data from large genome-wide association studies,²⁰ excessive fat in both limbs and trunk was shown to increase OSA risk. However, this study’s fat measurements were limited to trunk and limbs, and these findings were exclusively based on European-ancestry populations. In an exploratory ancillary study of individuals with sleep disorders, Tan et al. observed associations between FM of trunk and android regions with OSA. However, this research was restricted to overweight or obese male participants.²¹ Degache et al. found that specific regional fat distribution is a prognostic factor for the severity of OSA. However, this research was confined to between-group comparisons in elderly subjects aged >65 years.²² Notably, a longitudinal study by Martin et al. demonstrated no effect of central FM on OSA. The study participants were recruited from a cross-sectional community-based survey of a homogenous group of elderly subjects enrolled through strict inclusion criteria, and therefore, the findings may not be fully representative of clinical populations, requiring particular caution in result extrapolation.²³ Our results showed that both central and peripheral FMI are associated with OSA, and central fat has a more significant impact on OSA than peripheral fat. Central obesity not only increases intra-abdominal pressure, reduces lung volume, and raises the likelihood of upper airway collapse²⁴ but also leads to an increase in visceral fat, which secretes various inflammatory and adipokines, contributing to systemic inflammation and oxidative stress, affecting upper airway muscle activity, promoting fat tissue growth around the upper airway, thereby increasing the risk of OSA.^25,26

Beyond the general association between fat distribution and OSA, our findings further reveal significant gender-specific patterns in this relationship. To explore potential effect modification by gender, we conducted gender-stratified analyses. The results suggested that the association between fat distribution and OSA was more pronounced in males, with the direction and magnitude of effects aligning with the primary findings. However, for females, the relationship between leg and gynoid FMI and OSA did not reach statistical significance. This discrepancy may originate from sexually dimorphic adiposity distribution patterns. Specifically, females tend to distribute fat around the buttocks and thighs, whereas males typically accumulate excess fat more centrally in the abdominal area.²⁷ This correlates with higher insulin resistance in males than in females.²⁸ However, this pattern changes after menopause in females, where fat storage becomes more centralized, and metabolic risks become more similar to those of males.²⁹ In addition, our results suggest that regardless of gender, an increase in central fat is a significant contributing factor to the pathogenesis of OSA, a finding also confirmed by a clinical cohort study of 96 patients with OSA (M/F = 60/36).³⁰

Previous studies have rarely addressed the differences in OSA symptoms and regional fat across ethnicities. Our ethnicity-stratified analyses revealed differential adiposity–OSA associations across racial/ethnic groups: In non-Hispanic Whites, there was a clear association between OSA and the FMI in the trunk, arm, leg, android, gynoid, and abdominal areas. While comparable effect sizes emerged for trunk/android/gynoid FMI in other Hispanics, the specific patterns of fat distribution may differ from those of non-Hispanic Whites. However, no statistically significant associations were detected in Mexican Americans or non-Hispanic Blacks. Genetic differences among ethnicities may lead to different patterns of fat accumulation and distribution,^29,31 which could affect the pathogenesis of OSA. In addition, lifestyle, dietary habits, and socioeconomic status may vary among ethnic groups,^32,33 potentially influencing fat distribution and its impact on OSA.

Our BMI-stratified analyses revealed a critical paradox: while regional adiposity universally predicted OSA risk across BMI categories, the magnitude exhibited profound obesity-status dependency. Notably, nonobese individuals exhibited markedly stronger associations between abdominal/android fat accumulation and OSA risk compared with obese counterparts, with the magnitude of association nearly doubling across key fat compartments, suggesting that regional fat distribution could be a more critical determinant of OSA risk than overall adiposity in nonobese populations. We posit that this phenomenon may stem from limited adipose expandability in nonobese individuals, leading to preferential deposition of visceral/ectopic fat that exerts disproportionate mechanical and metabolic effects. Supporting this perspective, a cross-sectional study by Lin et al. found that higher FM in the trunk, calf, and arms was associated with increased risks of metabolic abnormalities, particularly in nonobese populations not classified as obese by BMI. This underscores the importance of assessing fat distribution beyond BMI for preventing metabolic syndrome.³⁴

This study has certain advantages, as body fat was measured using the gold-standard DXA, and as an important epidemiological study in the United States, NHANES has strict quality control procedures for data collection, making the conclusions quite reliable. In addition, the study adjusted for potential sociodemographic and lifestyle confounders to achieve more robust results. We used a weighted sample analysis to test the associations, making the results more generalizable. However, this study still has the following limitations: (1) Because of its cross-sectional design, it cannot observe the dynamic changes in body indices. (2) DXA screening was only conducted on adolescents and middle-aged adults, so the results cannot be generalized to older adults. (3) We diagnosed OSA based on some typical symptoms found in the NHANES questionnaire, such as daytime sleepiness, apnea, and snoring. Recall or self-report bias becomes a limitation of using questionnaires to collect information, but using the NHANES database, there are multiple studies on OSA.^35
–37

This large-scale population study demonstrates that regional fat distribution, particularly abdominal adiposity, is independently associated with OSA risk, with the abdominal FMI exhibiting superior predictive capacity across demographic subgroups. The differential associations observed by gender (male-specific universal associations vs. female-selective patterns), race (strongest in non-Hispanic Whites), and obesity status (enhanced effects in nonobese individuals) collectively suggest that OSA pathogenesis involves fat compartment-specific mechanisms interacting with population characteristics. Crucially, the identification of abdominal fat accumulation as a consistent risk determinant, even in nonobese individuals, underscores its potential utility as a screening biomarker for OSA susceptibility beyond conventional BMI thresholds. These findings advocate for incorporating regional adiposity assessment into OSA risk stratification frameworks, particularly in populations where traditional obesity metrics may fail to capture true metabolic risk.

Footnotes

Acknowledgments

We utilized data from the NHANES database. We gratefully acknowledge the National Center for Health Statistics for providing this public resource.

Authors’ Contributions

T.W. and G.Z. proposed the theme of this study, performed methodology and formal analysis, completed the writing of the original draft. L.T., Y.O., and H.L. were responsible for investigation and validation and also coordinated all works. X.Z. and Y.C. conducted data collection and visualization. J.P. supervised and checked the article writing and arranged funding acquisition. All authors are responsible for all aspects of the study, ensuring that any issues related to the accuracy and completeness of any part of the work are appropriately investigated and addressed.

Data Availability

The datasets generated and analyzed for the current study are available in the NHANES repository. These data can be obtained from the following link: .

Ethical Approval

The ethics review for this study was approved by the National Center for Health Statistics Institutional Review Board. The methods used in this study were conducted in accordance with relevant guidelines and regulations, including the Declaration of Helsinki. All individuals provided written informed consent prior to participating in the study. For more details, please refer to .

Author Disclosure Statement

No conflicting financial interests exist.

Funding Information

This work was supported by grants from the National Key Research and Development Program of China (grant no. 2022YFC2503902).

References

Gottlieb

, Punjabi

. Diagnosis and management of obstructive sleep apnea: A review. JAMA, 2020; 323(14):1389–1400; doi: 10.1001/jama.2020.3514

Javaheri

, Javaheri

. Update on persistent excessive daytime sleepiness in OSA. Chest, 2020; 158(2):776–786; doi: 10.1016/j.chest.2020.02.036

Yeghiazarians

, Jneid

, Tietjens

, et al. Obstructive sleep apnea and cardiovascular disease: A scientific statement from the American Heart Association. Circulation, 2021; 144(3):e56–e67; doi: 10.1161/cir.0000000000000988

Paschou

, Bletsa

, Saltiki

, et al. Sleep apnea and cardiovascular risk in patients with prediabetes and Type 2 diabetes. Nutrients, 2022; 14(23); doi: 10.3390/nu14234989

Benjafield

, Ayas

, Eastwood

, et al. Estimation of the global prevalence and burden of obstructive sleep apnoea: A literature-based analysis. Lancet Respir Med, 2019; 7(8):687–698; doi: 10.1016/s2213-2600(19)30198-5

Jennum

, Kjellberg

. Health, social and economical consequences of sleep-disordered breathing: A controlled national study. Thorax, 2011; 66(7):560–566; doi: 10.1136/thx.2010.143958

Jennum

, Ibsen

, Kjellberg

. Social consequences of sleep disordered breathing on patients and their partners: A controlled national study. Eur Respir J, 2014; 43(1):134–144; doi: 10.1183/09031936.00169212

Tregear

, Reston

, Schoelles

, et al. Obstructive sleep apnea and risk of motor vehicle crash: Systematic review and meta-analysis. J Clin Sleep Med, 2009; 5(6):573–581.

Isono

. Obesity and obstructive sleep apnoea: Mechanisms for increased collapsibility of the passive pharyngeal airway. Respirology, 2012; 17(1):32–42; doi: 10.1111/j.1440-1843.2011.02093.x

10.

Babb

, Wyrick

, DeLorey

, et al. Fat distribution and end-expiratory lung volume in lean and obese men and women. Chest, 2008; 134(4):704–711; doi: 10.1378/chest.07-1728

11.

Welch

, Foster

, Ritter

, et al. A novel volumetric magnetic resonance imaging paradigm to study upper airway anatomy. Sleep, 2002; 25(5):530–540.

12.

Seidell

. Waist circumference and waist/hip ratio in relation to all-cause mortality, cancer and sleep apnea. Eur J Clin Nutr, 2010; 64(1):35–41; doi: 10.1038/ejcn.2009.71

13.

Davies

, Ali

, Stradling

. Neck circumference and other clinical features in the diagnosis of the obstructive sleep apnoea syndrome. Thorax, 1992; 47(2):101–105; doi: 10.1136/thx.47.2.101

14.

Bell

, Carslake

, O’Keeffe

, et al. Associations of body mass and fat indexes with cardiometabolic traits. J Am Coll Cardiol, 2018; 72(24):3142–3154; doi: 10.1016/j.jacc.2018.09.066

15.

Jabłonowska-Lietz

, Wrzosek

, Włodarczyk

, et al. New indexes of body fat distribution, visceral adiposity index, body adiposity index, waist-to-height ratio, and metabolic disturbances in the obese. Kardiol Pol, 2017; 75(11):1185–1191; doi: 10.5603/KP.a2017.0149

16.

Kelly

, Wilson

, Heymsfield

. Dual energy X-Ray absorptiometry body composition reference values from NHANES. PLoS One, 2009; 4(9):e7038; doi: 10.1371/journal.pone.0007038

17.

Cavallino

, Rankin

, Popescu

, et al. Antimony and sleep health outcomes: NHANES 2009-2016. Sleep Health, 2022; 8(4):373–379; doi: 10.1016/j.sleh.2022.05.005

18.

Zhou

, Chen

, Mao

, et al. Association between obstructive sleep apnea and visceral adiposity index and lipid accumulation product: NHANES 2015-2018. Lipids Health Dis, 2024; 23(1):100; doi: 10.1186/s12944-024-02081-5

19.

, Zhao

, Yu

, et al. Short sleep duration was associated with increased regional body fat in US Adults: The NHANES from 2011 to 2018. Nutrients, 2022; 14(14); doi: 10.3390/nu14142840

20.

Sun

, Zeng

, Gao

, et al. Causal associations between sarcopenia-related traits and obstructive sleep apnea: A Mendelian randomization study. Aging Clin Exp Res, 2025; 37(1):68; doi: 10.1007/s40520-025-02963-3

21.

Tan

, Alén

, Cheng

, et al. Associations of disordered sleep with body fat distribution, physical activity and diet among overweight middle-aged men. J Sleep Res, 2015; 24(4):414–424; doi: 10.1111/jsr.12283

22.

Degache

, Sforza

, Dauphinot

, et al.; PROOF Study Group. Relation of central fat mass to obstructive sleep apnea in the elderly. Sleep, 2013; 36(4):501–507; doi: 10.5665/sleep.2532

23.

Saint Martin

, Roche

, Thomas

, et al. Association of body fat composition and obstructive sleep apnea in the elderly: A longitudinal study. Obesity (Silver Spring), 2015; 23(7):1511–1516; doi: 10.1002/oby.21121

24.

Harada

, Oga

, Chihara

, et al. Differences in associations between visceral fat accumulation and obstructive sleep apnea by sex. Ann Am Thorac Soc, 2014; 11(3):383–391; doi: 10.1513/AnnalsATS.201306-182OC

25.

Senaratna

, Perret

, Lodge

, 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

26.

Veasey

, Rosen

. Obstructive sleep apnea in adults. N Engl J Med, 2019; 380(15):1442–1449; doi: 10.1056/NEJMcp1816152

27.

Whittle

, Marshall

, Mortimore

, et al. Neck soft tissue and fat distribution: Comparison between normal men and women by magnetic resonance imaging. Thorax, 1999; 54(4):323–328; doi: 10.1136/thx.54.4.323

28.

Geer

, Shen

. Gender differences in insulin resistance, body composition, and energy balance. Gend Med, 2009; 6(Suppl 1):60–75; doi: 10.1016/j.genm.2009.02.002

29.

Guglielmi

, Sbraccia

. Obesity phenotypes: Depot-differences in adipose tissue and their clinical implications. Eat Weight Disord, 2018; 23(1):3–14; doi: 10.1007/s40519-017-0467-9

30.

Simpson

, Mukherjee

, Cooper

, et al. Sex differences in the association of regional fat distribution with the severity of obstructive sleep apnea. Sleep, 2010; 33(4):467–474; doi: 10.1093/sleep/33.4.467

31.

Lee

, Wu

, Fried

. Adipose tissue heterogeneity: Implication of depot differences in adipose tissue for obesity complications. Mol Aspects Med, 2013; 34(1):1–11; doi: 10.1016/j.mam.2012.10.001

32.

Ralls

, Grigg-Damberger

. Roles of gender, age, race/ethnicity, and residential socioeconomics in obstructive sleep apnea syndromes. Curr Opin Pulm Med, 2012; 18(6):568–573; doi: 10.1097/MCP.0b013e328358be05

33.

Goosmann

, Williams

, Springer

, et al. The impact of marital status and race in obstructive sleep apnea. Ear Nose Throat J, 2022:1455613221120068; doi: 10.1177/01455613221120068

34.

Lin

, Hu

, He

, et al. Regional adipose distribution and metabolically unhealthy phenotype in Chinese adults: Evidence from China National Health Survey. Environ Health Prev Med, 2025; 30:5; doi: 10.1265/ehpm.24-00154

35.

Wang

, Shi

, Lin

, et al. Association between the triglyceride glucose index and obstructive sleep apnea and its symptoms: Results from the NHANES. Lipids Health Dis, 2024; 23(1):133; doi: 10.1186/s12944-024-02125-w

36.

Kadier

, Dilixiati

, Ainiwaer

, et al. Analysis of the relationship between sleep-related disorder and systemic immune-inflammation index in the US population. BMC Psychiatry, 2023; 23(1):773; doi: 10.1186/s12888-023-05286-7

37.

Tao

, Niu

, Lu

, et al. Obstructive sleep apnea (OSA) is associated with increased risk of early-onset sarcopenia and sarcopenic obesity: Results from NHANES 2015-2018. Int J Obes (Lond), 2024; 48(6):891–899; doi: 10.1038/s41366-024-01493-8

Association Between Obstructive Sleep Apnea and Regional Fat: A Cross-Sectional Analysis of National Health and Nutrition Examination Survey 2015–2018