The association between cardiovascular health, measured using Life’s Essential 8, and obstructive sleep apnea: a cross-sectional and Mendelian randomization study

Abstract

Background: Cardiovascular health (CVH), operationalized using the Life’s Essential 8 (LE8) construct, has been linked to a myriad of health outcomes. Although there is a link between CVH and Obstructive Sleep Apnea (OSA), which has become a significant public health issue, a thorough investigation of this relationships has not been undertaken. Methods: This study investigates the association between CVH, quantified according to the LE8 construct, and OSA using data from the National Health and Nutrition Examination Survey (NHANES). Based on data from the NHANES (2005-2008, 2015-2018), this study consisted of 9,350 adults aged 20-79 years. Logistic regression analysis is utilized to determine the association between LE8 scores and OSA. Two-sample Mendelian randomization (MR) is used to investigate the causal relationship between CVH and OSA with the inverse-variance weight (IVW) method being the primary source of analysis. Results: The connection between CVH and OSA risk was affirmed, with individuals who had better CVH (high LE8 vs. low LE8), being at significantly lower risk for OSA than those with worse CVH (Odds Ratio [OR]: 0.23, 95% confidence interval [CI]: 0.18-0.30). Additionally, patients with higher LE8 scores experienced a decreased risk of OSA in a manner that was dose-dependent. The MR analysis also provided evidence of a positive correlation between BMI (genetically determined) and increased risk for OSA (OR for 1 Standard Deviation [SD] increase = 2, 95% CI: 1.81-2.21). Additionally, individuals who smoked less and exercised more showed decreased risk of OSA with respect to smoking (OR = 0.5, 95% CI: 0.32-0.79) and exercise (OR = 0.73, 95% CI: 0.54-0.99). Conclusions: This study highlights the importance of keeping CVH in good shape (as measured by the LE8 score) to lower the risk of developing OSA.

Introduction

There are many ways to identify obstructive sleep apnea (OSA) which is complex in nature, yet a highly common condition, where a person will have repeated interruptions in their breathing or shallow breathing while they sleep [1,2]. OSA can be diagnosed using multiple techniques and may be associated with a number of different cardiovascular diseases (CVD) [3]. Studies have shown that many people with OSA go undiagnosed, even though it affects a large number of people [4]. Approximately 13%-33% of adult men and approximately 6%-19% of adult women within the United States have OSA [5-7], which represents almost 1 billion people around the world [8]. Society also incurs significant costs associated with OSA, beginning from the time before an individual is diagnosed with OSA when they will lose out on potential earnings, to the time after they have been successfully treated for OSA. This is indicative of the need for improved awareness and early detection of OSA, as well as more effective management practices for improved health and reduced healthcare costs associated with OSA [9].

There is strong evidence that OSA is an independent risk factor for hypertension [10], diabetes mellitus [11], and CVD [3]. In 2010, Life’s Simple 7 (LS7) [12] introduced by the American Heart Association (AHA) identified seven important factors affecting heart health: Diet, Smoking, Body Weight (BMI), Physical Activity, Blood Pressure, Fasting Blood Sugar [13,14] and Cholesterol Levels. Some of these factors, including Blood Pressure, Fasting Blood Glucose [15] and BMI [16], have a two-way relationship with OSA, demonstrating that the impact of one condition can affect another and vice versa. Subsequent studies have demonstrated the presence of a strong association between OSA and LS7 scores [17]. In 2022, the AHA expanded upon LS7 with Life’s Essential 8 (LE8) [1] to promote further improvement in heart health. The higher one’s LE8 scores, the better the health outcomes (e.g., Reduced CVD Rates, Reduced Rates of Other Major Illnesses, Reduced Rates of Total Mortality) [18-20]. Given the close link between OSA and cardiovascular health, strategies that improve components of the LE8 construct may play an important role in reducing OSA risk. However, the connection between LE8 scores and OSA needs to be thoroughly researched, which provides an opportunity for future study.

In the current research, we examined substantial datasets from the National Health and Nutrition Examination Surveys (NHANES) and Genome-Wide Association Studies (GWAS) performed by the United Kingdom Biobank (UKB). A primary focus was assessing how LE8 score levels at baseline (based on health behaviours and health risk factors, etc.) relate to the development of OSA. We also examined both linear and nonlinear associations of LE8 score levels with the incidence of OSA. Finally, we employed Mendelian Randomization (MR) to establish a cause-and-effect relationship between LE8 score levels and the development of OSA.

Method and materials

Study design and population in NHANES

For this investigation we utilized data from the National Health and NHANES, collected by the Centers for Disease Control and Prevention (CDC) and National Center for Health Statistics (NCHS) for the years 2005-2008 and 2015-2018. The NHANES methodology is based on a stratified, multistage method of sampling a representative population of the United States who are not currently institutionalized. In order to ensure that we provided a complete and comprehensive account of our research, we followed the Strengthening of Reporting of Observational Studies in Epidemiology (STROBE) guiding principles.

Adults aged 20 years or older were included from the NHANNES surveys (2005-2008 and 20152018). Our inclusion criteria were purposely broad to include as many people as possible. We excluded a large number of participants for a variety of reasons. For example, participants who were missing demographic information (age, gender, race/ethnic group, poverty income ratio (PIR), pregnant women (n=518), did not have data on Obstructive Sleep Apnea (2,047 individuals), missing LE8 data (7,408), and missing CVD data (2 individuals), etc. Exclusions a total of 30,372 individuals, providing us with 9,350 subjects for the analysis (Figure 1). This was done in order to have a quality dataset to work from.

Figure 1.

Selection of participants in the study. OSA, obstructive sleep apnea; NHANES, national health and nutrition examination survey; CVD, cardiovascular disease.

Measurement of LE8 in HNANES

In this study, we measured LE8 based on the American Heart Association’s Presidential Advisory guidelines [2], as shown in Table S1. The LE8 construct is used to operationalize cardiovascular health according to AHA guidance by considering four health behaviors - diet, physical activity, tobacco use, and sleep - and four health factors - BMI, non-HDL cholesterol, blood glucose, and blood pressure. Each of these indicators is scored from 0 to 100, with higher scores meaning better health [1]. To calculate the total LE8 score for an individual, we take the mean of the eight component scores; therefore, an individual can have a total LE8 score that has both a continuous and categorical value associated with it based on the AHA-defined cut-off values of 50 and 79 to determine CVH categories of low, moderate, or high CVH results based on the mean scoring of all eight components [1]. As an example, for assessing diet we used the 2015 Healthy Eating Index (HEI) [21], with scoring information in Table S2, and blood glucose control was assessed via glycated hemoglobin (HbA1c) levels as part of the LE8 score assessment process.

Measurements and definition of OSA in NHANES

This research was conducted using NHANES data, with the objective of clarifying what we meant by OSA and diagnosing it accurately. We determined whether or not individuals were diagnosed with OSA based on their answers to three distinct questions concerning common symptoms associated with OSA [22]. OSA was defined as meeting any of the following criteria: (1) Frequency of snoring - occurs at least 3 times per week; (2) Witnessed apneas - happening at least 3 times per week; (3) Excessive daytime sleepiness - experienced 16-30 times per month, when getting an average of 7 hours or more of sleep every night during the week and/or while working.

Mendelian randomization

Our analysis of the interplay of modifiable risk factors in Life’s Essential 8 with OSA involved performing an extensive two-sample MR analysis [23]. Using SNPs as instruments and analyzing both genetic and environmental information, we tested the risk factors with the appropriate SNPs [24]. We established that our results were robust by following the three assumptions of MR: (1) SNPs must strongly associate with the risk factors, (2) SNPs must not be confounded by any external issues that would alter the exposure-outcome relationship, and (3) all manipulations from SNPs. We used selected genetic variants as instruments to perform our analyses, as previously performed by UKB, for this study.

Detailed descriptions of how these genetic instruments were created and applied can be reviewed in the Supplementary Information sections with complete lists shown in Tables S3, S4, S5, S6, S7, S8, S9, S10. The GWAS on which our analyses were based on OSA was obtained from the Finnish National Genome Project (FinnGen), following protocols established by the Global Alliance for Genomics and Health. This database includes a total of 217,955 people, with 16,761 OSA cases and 201,194 controls, for an available set of 16,380,465 SNPs for analysis. The diagnosis of OSA was made using ICD-10 code G47.3 and ICD-9 code 3472A and based on clinical diagnosis of symptoms and additionally confirmed using clinical exams and polysomnography [25].

Statistical analysis in a cross-sectional study

Using sample weights, clustering and stratification allowed us to make sure that our estimates were representative of the total U.S. population based on the complicated sampling designs used in NHANES. Categorical data presented as weighted percentages and continuous data presented as averages, with 95% CIs providing an indicator of the accuracy of our results. To compare the characteristics of NHANES participants, we used Rao-Scott chi-square testing and linear regression analysis without adjustment to provide a complete picture of the demographic and clinical factors related to CVH and to calculate unadjusted rates of CVH categories based on the overall population.

Using survey weighted logistic regression models allowed us to evaluate the relationship between OSA and LE8 Scores. Models 1-3 provide different levels of adjustment based on multiple covariates; Age, Gender, Race & Ethnicity in Model 1; Socioeconomic Variables (Poverty Ratio, Education and Marital Status) in Model 2; CVD and Obesity in Model 3. In addition, restricted cubic splines helped determine possible non-linear relationships between LE8 scores (and their individual components) and OSA, and we conducted likelihood ratio tests to evaluate those relationships for non-linearity.

Also notable is that we performed subgroup analyses concerning demographic characteristics of each study participant to assess if CVH has different effects on OSA in each of these subgroups, including exploration of interaction terms to assess how these groups interact collectively. Statistical tests were conducted bilaterally (two-sided) and were computed using R software, Version 4.2.1. All statistical results were thought to be statistically significant when P values were less than 0.05.

Statistical analysis in Mendelian randomization study

The main analysis was conducted using the IVW method with a random effects model to account for heterogeneity (or variances) among genetic tools used in the analyses [26]. To strengthen the findings from the IVW analysis further, we also performed additional analysis using both MR-Egger and weighted median methods [27]. To check that our genetic instruments were consistent, we tested for SNP variability using Cochran’s Q tests [28]. In addition to these methods, we further verified by ruling out the effect of directional pleiotropy from genetic instruments, we conducted an MR-Egger regression where we examined the intercept value in order to assess potential confounding factors associated with genetic risk factors that might influence health through alternate pathways not assessed in this study. We utilized Cochran’s Q statistic to estimate the degree of heterogeneity in the IVW MR framework and defined a statistically significant finding for each SNP to be any resultant Cochran’s Q statistic with a P value less than 0.05.

Results

Baseline characteristics

The study sample consisted of 9,350 individuals aged 20 years and older who were separated by OSA status based on the demographic results presented in Table 1. The mean age of the participants was 47.75 years (Standard Error [SE]: 0.37), of which 52.11% were female (weighted percentage). The mean LE8 score across all participants was 67.81 (95% CI: 67.39-68.23), with the percentage of CVH levels classified as high being 34.84%, moderate being 50.48%, and low being 14.68%. Approximately 29.1% of individuals in the sample were diagnosed with OSA, resulting in a total of 2,993 participants with a diagnosis of OSA.

Table 1.

Baseline characteristics of the study population

	Overall (n = 10300)	Non OSA (n = 7307)	OSA (n = 2993)	P value
Age, years	47.75±0.37	47.47±0.40	48.40±0.45	0.03
Age strata				<0.0001
20-39	3291 (34.21)	2469 (36.22)	822 (29.48)
40-59	3449 (40.36)	2253 (37.62)	1196 (46.83)
≥60	3560 (25.43)	2585 (26.16)	975 (23.69)
sex				<0.0001
Female	5249 (52.11)	3900 (55.41)	1349 (44.30)
Male	5051 (47.89)	3407 (44.59)	1644 (55.70)
Race/ethnicity				0.19
Black	2056 (9.79)	1464 (9.81)	592 (9.74)
White	4926 (72.87)	3470 (72.49)	1456 (73.79)
Mexican	1738 (7.27)	1268 (7.62)	470 (6.44)
Others	1580 (10.06)	1105 (10.08)	475 (10.03)
Poverty ratio				0.57
<1.3	4101 (36.48)	2889 (36.27)	1212 (36.98)
1.3-3.5	2818 (17.30)	2038 (17.58)	780 (16.65)
>3.5	3381 (46.22)	2380 (46.15)	1001 (46.37)
Education levels				<0.0001
High school or less	4950 (38.44)	3488 (37.31)	1462 (41.12)
Some college or associates degree	2952 (31.43)	2044 (30.85)	908 (32.77)
College graduate or above	2398 (30.13)	1775 (31.84)	623 (26.10)
Marital status				<0.0001
Coupled	7886 (80.28)	5472 (78.58)	2414 (84.31)
Single or separated	2414 (19.72)	1835 (21.42)	579 (15.69)
LE8 scores (out of 100 possible points)	67.81±0.42	69.71±0.41	63.31±0.51	<0.0001
HEI-2015 diet score	38.96±0.81	40.06±0.86	36.35±0.86	<0.0001
Physical activity score	68.74±0.79	69.39±0.92	67.18±1.17	0.11
Nicotine exposure score	70.16±0.79	72.25±0.80	65.23±1.14	<0.0001
Sleep health score	84.02±0.46	85.57±0.44	80.35±0.76	<0.0001
Body mass index score	61.25±0.66	66.13±0.66	49.72±0.91	<0.0001
Blood lipids score	63.49±0.37	64.98±0.44	59.99±0.68	<0.0001
Blood glucose score	86.91±0.40	88.34±0.44	83.54±0.62	<0.0001
Blood pressure score	68.93±0.45	70.95±0.48	64.14±0.68	<0.0001
CVH				<0.0001
High	3162 (34.84)	2559 (40.01)	603 (22.61)
Moderate	5319 (50.48)	3661 (47.92)	1658 (56.54)
Low	1819 (14.68)	1087 (12.07)	732 (20.85)
Health behaviors score	65.47±0.59	66.82±0.59	62.28±0.74	<0.0001
Health factors score	70.15±0.33	72.60±0.34	64.35±0.44	<0.0001
CVD status				<0.001
No	9131 (91.71)	6547 (92.52)	2584 (89.79)
Yes	1169 (8.29)	760 (7.48)	409 (10.21)
Obesity status				<0.0001
No	6425 (63.78)	4947 (69.85)	1478 (49.43)
Yes	3875 (36.22)	2360 (30.15)	1515 (50.57)

Low CVH was defined as a LE8 score of 0 to 49, moderate CVH of 50-79, and high CVH of 80-100. LE8, Life’s Essential 8; CVH, Cardiovascular health; CVD, cardiovascular disease.

In comparing both groups together, we see that the individuals who had OSA were predominantly male, had a lower educational level than those without OSA, were more often in supportive relationships, and had a higher prevalence of cardiovascular disease and obesity when compared to those who did not suffer from OSA. However, despite these differences, there was no significant difference in physical activity between the two groups, as shown in Table 1.

Higher LE8 scores are associated with a lower prevalence of OSA

After adjusting for age, we found a clear negative relationship between OSA prevalence and CVH levels, as measured by the LE8 score. The results showed a big difference in OSA rates across the CVH categories: people with high CVH had a much lower OSA rate of 16.2% (95% CI: 14.6-17.8), compared to 31.1% (95% CI: 29.8-32.4) in the moderate CVH group and 48.4% (95% CI: 43.7-53.0) in the low CVH group (Figure 2).

Figure 2.

Age-adjusted prevalence of OSA in different levels of Life’s Essential 8 scores. Numbers at the top of the bars represent the weighted percentage. Bar whiskers represent the 95% confidence level. Low CVH was defined as a LE8 score of 0 to 49, moderate CVH of 50-79, and high CVH of 80-100.

Our analysis confirmed that people with better cardiovascular health (CVH) were much less likely to have OSA. Those with moderate CVH had an Odds Ratio (OR) of 0.52 (95% CI: 0.45-0.63), and those with high CVH had an OR of 0.23 (95% CI: 0.18-0.30), compared to people with low CVH (Table 2). When we looked at how the LE8 score impacted OSA, we found that for every 10-point increase in the LE8 score, the chances of having OSA went down significantly (OR: 0.72, 95% CI: 0.68-0.76), as shown in Table 2. We also saw a clear pattern where better CVH (higher LE8 scores) was linked to lower OSA rates, as shown in Figure 3A.

Table 2.

Association of the Life’s Essential 8 scores with obstructive sleep apnea (OSA)

	Unadjusted model		Multivariable model 1*		Multivariable model 2†		Multivariable model 3‡
	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value	OR (95% CI)	p value
CVH
Low (0-49)	ref		ref		ref		ref
Moderate (50-79)	0.53 (0.46, 0.62)	<0.01	0.52 (0.44, 0.61)	<0.01	0.51 (0.43, 0.61)	<0.01	0.52 (0.44, 0.63)	<0.01
High (80-100)	0.22 (0.18, 0.27)	<0.01	0.22 (0.18, 0.28)	<0.01	0.23 (0.18, 0.29)	<0.01	0.23 (0.18, 0.30)	<0.01
Per 10 points increase	0.72 (0.69, 0.75)	<0.01	0.72 (0.69, 0.75)	<0.01	0.72 (0.68, 0.76)	<0.01	0.72 (0.68, 0.76)	<0.01
Health behaviors score
Low (0-49)	ref		ref		ref		ref
Moderate (50-79)	0.72 (0.62, 0.83)	<0.01	0.71 (0.61, 0.83)	<0.01	0.72 (0.61, 0.85)	<0.01	0.73 (0.62, 0.86)	<0.01
High (80-100)	0.55 (0.47, 0.65)	<0.01	0.55 (0.46, 0.65)	<0.01	0.58 (0.47, 0.71)	<0.01	0.59 (0.48, 0.72)	<0.01
Per 10 points increase	0.89 (0.86, 0.91)	<0.01	0.89 (0.86, 0.92)	<0.01	0.89 (0.86, 0.93)	<0.01	0.89 (0.86, 0.93)	<0.01
Health factors score
Low (0-49)	ref		ref		ref		ref
Moderate (50-79)	0.68 (0.58, 0.80)	<0.01	0.64 (0.55, 0.76)	<0.01	0.65 (0.55, 0.77)	<0.01	0.66 (0.56, 0.78)	<0.01
High (80-100)	0.33 (0.28, 0.38)	<0.01	0.31 (0.27, 0.36)	<0.01	0.32 (0.28, 0.37)	<0.01	0.33 (0.28, 0.38)	<0.01
Per 10 points increase	0.79 (0.77, 0.81)	<0.01	0.78 (0.76, 0.80)	<0.01	0.79 (0.76, 0.81)	<0.01	0.79 (0.76, 0.81)	<0.01

CVH, Cardiovascular health; OR, odds ratio; CI, confidence interval.

^*Adjusted for age, sex, race/ethnicity.

^†Additionally adjusted for poverty ratio, education levels, and marital status.

^‡Additionally adjusted for CVD status, obesity status.

Figure 3.

Dose-response relationships between Life’s Essential 8 scores (A), Health Behavior score (B), Health Factors score (C), and OSA. ORs (solid lines) and 95% confidence levels (shaded areas) were adjusted for age, sex, race/ethnicity, obesity; poverty ratio (as a continuous variable), education level, marital status and CVD status. Vertical dotted lines indicate the minimal threshold for the beneficial association with estimated OR = 0. OR: odds ratio, LE8: Life’s Essential 8.

Higher health behavior scores are linked to reduced, but nonlinearly decreasing, OSA risk

After adjusting for age, we found that OSA rates were significantly lower in people with high health behavior scores (23.5%, 95% CI: 21.8-25.3%) compared to those with moderate (29.0%, 95% CI: 27.5-30.5%) and low health behavior scores (36.8%, 95% CI: 33.9-39.6%), as shown in Figure 2. The statistical analysis supports the concept that improved health behaviour choices directly correlate with decreased risk of developing OSA. Individuals with moderate and high health behaviour ratings were significantly less likely to have OSA than individuals with a poor health behaviour rating (OR: 0.73; 95% CI: 0.62-0.86 and OR: 0.59; 95% CI: 0.48-0.72, respectively) (Table 2). Every increase of 10 points in health behaviour rating was associated with a reduction in the chance of developing OSA. The odds of developing OSA decreased by 11% (OR: 0.89; 95% CI: 0.86-0.93) for every increase of 10 points in health behaviour rating (Table 2). We also found that the link between health behavior scores and OSA risk isn’t linear (P for non-linearity <0.001) (Figure 3B), meaning the risk of OSA drops less as health behavior scores go up.

Improved health factor scores are linked to a non-linearly decreasing risk of OSA

After adjusting for age, we saw a significant drop in OSA rates among people with higher health factor scores. Those with high health factors had an OSA rate of 19.2% (95% CI: 17.8-20.7), compared to 33.0% (95% CI: 31.3-34.8) for those with moderate health factors and 44.2% (95% CI: 40.6-47.8) for those with low health factors, as shown in Figure 2. After adjusting for other factors, we found that people with moderate and high health factor scores had much lower odds of having OSA - OR = 0.66 (95% CI: 0.56-0.78) and OR = 0.33 (95% CI: 0.28-0.38), respectively - compared to those with low health factors. Additionally, every 10-point increase in the health factor score was linked to a lower odds of having OSA (OR = 0.79, 95% CI: 0.76-0.81) (Table 2). We also found that the relationship between health factor scores and OSA risk wasn’t linear (non-linearity P = 0.002), meaning that as health factor scores improved, OSA risk kept dropping, but at a slower rate (Figure 3C).

Stronger LE8-OSA protective effect in younger and higher-educated groups. As shown in Figure 4, our subgroup analyses consistently found a negative relationship between the LE8 score and the likelihood of having OSA across different groups. A key finding was that there were significant interaction effects between the LE8 score and age as well as education levels in relation to OSA risk (interaction P < 0.05). Specifically, the link between higher LE8 scores and lower OSA risk was stronger in younger people (aged 20-39 years; OR per 10-point increase, 0.65; 95% CI: 0.59-0.71) and those with higher education, with some college or an associate degree (OR per 10-point increase, 0.68; 95% CI: 0.63-0.73).

Figure 4.

Subgroup analysis of the association of the Life’s Essential 8 scores and the presence of OSA. ORs were calculated as per 10 scores increase in LE8 score. Each stratification was adjusted for age, sex, race/ethnicity, poverty ratio, education level, and marital status. An OR >1 indicates increased risk while <1 indicates decreased risk. OR odds ratio, CI confidence interval.

MR identifies causal roles of BMI, smoking, and physical activity in OSA risk

Using two independent datasets, we performed MR analyses to determine whether modification of eight risk factors has a causal relationship with OSA. We examined the effect of genetic variants associated with each of the eight factors (SBP, LDL-C, HbA1c, smoking behaviour, BMI, level of physical activity, sleep duration and dietary intake). Each of these genetic variants enabled us to establish independent causal relationships between each modifiable risk factor and OSA risk. Specifically, our main IVW MR analysis identified that there is a strong causal relationship between a genetic predisposition to having a higher BMI and developing OSA (OR for a 1 SD increase = 2, 95% CI: 1.81-2.21, P<0.001) (Figure 5). Conversely, a genetic predisposition to smoking less or engaging in more physical activity significantly decreased the risk of developing OSA (ORs = 0.50, 95% CI: 0.32-0.79, P = 0.003; OR = 0.73, 95% CI: 0.54-0.99, P = 0.041, Figure 5). All of these findings were subsequently confirmed using other MR analytic methods including MR-Egger and the weighted median method (Table 3). Conversely, there was no evidence of an independent causal relationship between genetically determined SBP, LDL-C, HbA1c, sleep duration or diet on the risk of developing OSA for either IVW or additional MR analyses (OR for a 1 SD increase = 2, 95% CI: 1.81-2.21).

Figure 5.

Forest plot of Mendelian randomization estimates between genetic predisposition to individual items of Life’s Essential 8 (LE8) and risk of OSA. The figure showed the IVW estimates of significantly OSA-associated modifiable risk factor of LE8. The blue dots represent the IVW estimates, and the black bars represent the 95% confidence intervals of IVW estimates. The OR >1 indicates increased risk while <1 indicates decreased risk.

Table 3.

Mendelian randomization analysis estimates from each method of assessing the causal effect of modifiable risk factors on the risk of OSA

Exposure	Method	SNPs	ORs	95% CI	P	Heterogeneity		Pleiotropy
						Q	P	Intercept	SE	P
SBP	IVW	233	1.12	(0.97, 1.28)	0.13	386.58	<0.01
	MR-Egger	233	0.89	(0.58, 1.33)	0.54			0.004	0.004	0.23
	Weighted median	233	1.15	(0.97, 1.37)	0.10
LDL cholesterol	IVW	151	0.10	(0.94, 1.07)	0.98	169.83	0.13
	MR-Egger	151	0.98	(0.90, 1.07)	0.65			0.001	0.002	0.49
	Weighted median	151	0.98	(0.89, 1.08)	0.68
HBA1C	IVW	316	1.01	(0.93, 1.09)	0.81	495.54	<0.01
	MR-Egger	316	0.90	(0.78, 1.04)	0.15			0.004	0.002	0.06
	Weighted median	316	0.97	(0.87, 1.08)	0.56
Smoking index	IVW	72	0.50	(0.32, 0.79)	0.00	98.42	0.02
	MR-Egger	72	0.11	(0.02, 0.74)	0.03			0.014	0.008	0.11
	Weighted median	72	0.55	(0.31, 0.97)	0.04
BMI	IVW	385	2.00	(1.81, 2.21)	<0.01	604.57	<0.01
	MR-Egger	385	2.15	(1.63, 2.83)	<0.01			-0.001	0.003	0.58
	Weighted median	385	2.06	(1.80, 2.36)	<0.01
Physical activity	IVW	11	0.73	(0.55, 0.99)	0.04	10.78	<0.01
	MR-Egger	11	0.92	(0.08, 11.16)	0.95			-0.007	0.036	0.86
	Weighted median	11	0.65	(0.44, 0.96)	0.03
Sleep duration	IVW	15	1.23	(0.84, 1.80)	0.28	32.63	<0.01
	MR-Egger	15	2.40	(0.65, 8.96)	0.21			-0.019	0.018	0.32
	Weighted median	15	1.31	(0.89, 1.93)	0.16
Diet score	IVW	68	0.83	(0.50, 1.35)	0.45	193.75	<0.01
	MR-Egger	68	0.67	(0.10, 4.80)	0.70			0.002	0.012	0.84
	Weighted median	68	0.74	(0.47, 1.18)	0.20

Discussion

This large nationwide study demonstrates that as LE8 scores increase, the likelihood of having OSA decreases. It appears from the data that both the health behaviours as well as the health factors included in the LE8 score are inversely associated with the risk of OSA. These findings were also demonstrated by subgroup analyses and showed the same trend across multiple subgroupings. This is among the first studies to demonstrate a clear dose-response relationship between CVH, defined using the LE8 construct, and OSA. The data also indicates that improvement in CVH leads to a continual decline in OSA risk, making it likely that adopting the behaviours necessary to increase LE8 scores would decrease your risk of developing OSA. Additionally, the data indicates a genetic relationship between increased BMIs and decreased levels of physical activity, both of which influence increased risk for OSA; therefore, it is possible that higher BMI and less exercise, both of which can cause decreased levels of fitness, can potentially lead to the development of OSA.

Previous studies have shown that health behaviors and factors like physical activity, diet, and BMI are often linked to OSA [29,30]. One study found an inverse relationship between OSA and the LS7 score [17], which is similar to our findings about the connection between LE8 scores and OSA. Therefore, early identification of high-risk OSA individuals is critical for their timely diagnosis and treatment. While obesity management has traditionally been listed among the recommendations to prevent OSA, our findings suggest a different focus for future prevention strategies [1,2,31,32]. Physical activity, which is an integral part of both preventing and treating obesity, has also demonstrated positive effects on OSA risk [30,33]. While the presence of a concurrent association between OSA, elevated blood pressure, and HbA1c levels has been previously described [13,15], our MR analysis did not reveal any evidence that increased risk of OSA from either elevated blood pressure or HbA1c levels could be attributed to a genetic component. This may indicate a possibility of reverse causality in observational studies. Another interesting outcome of our MR analysis is that there is a relationship between lower OSA risk and smoking. Further investigation into how smoking is related to both OSA and Sleep Health is required [34], more research is needed to better understand how smoking relates to sleep health.

Increasing evidence connects Lifestyle Environments 8 with Obstructive Sleep Apnea (OSA), primarily mediated by the lifestyle variables of Physical Activity (PA) and Body Mass Index (BMI) that are included in the definitions of LE8. Belly fat, in particular, contributes to an increased incidence, or severity, of OSA by collapsing the throat; thus, causing multiple instances of airway obstruction while sleeping. Mechanical pressure and nerve/muscle pathways are the two primary means via which belly fat can collapse the throat [35]. Furthermore, belly fat creates additional mechanical pressure in the airway and impairs the body’s ability to respond to that pressure, potentially through the influence of fat-related hormones (e.g., adipokines) on the nervous system. In summary, while belly fat can lead to greater incidence or severity of OSA, physical activity improves weight loss and also supports improved cardiovascular and metabolic health across a wide variety of disorders [32].

Despite some limitations within the study’s design, methodology, and reporting; the strengths of the research included various important components for each of the research objectives. First, this research used an extensive sample from the National Health and NHANES that included a diversity of individuals thus capturing representative information regarding the prevention and treatment of OSA. Second, the use of MR allowed researchers to determine if a causal relationship existed between the LE8 score and OSA. While researchers believe they have identified a statistically significant association between LE8 score and OSA, they must also consider that the results could be due in part to bias introduced from using self-reported health behaviours questionnaires (i.e., lifestyle choices). Lastly, since the majority of the NHANES study population was American, researchers cannot assume that the same findings would hold true for individuals of other ethnicities or demographics.

Conclusion

This research establishes an association between CVH, defined using the LE8 construct, and OSA. The findings highlight the importance of evaluating CVH among individuals with OSA, as lower levels of LE8-defined cardiovascular health were associated with a higher risk of OSA. The MR analysis demonstrated that weight management and physical activity are important strategies for managing OSA and to illustrate the importance of lifestyle changes in reducing the risk of OSA.

Disclosure of conflict of interest

None.

Abbreviations

CVH

Cardiovascular health

LE8

Life’s Essential

OSA

Obstructive sleep apnea

NHANES

National Health and Nutrition Examination Surveys

MR

Mendelian Randomization

IVW

inverse variance weighted

CVD

Cardiovascular disease

LS7

Life’s Simple 7

AHA

American Heart Association

BMI

Body mass index

GWAS

Genome-Wide Association Studies

UKB

United Kingdom Biobank

CDC

Centers for Disease Control and Prevention

NCHS

National Center for Health Statistics

HEI

Healthy Eating Index

HbA1c

glycated hemoglobin

SNP

Single Nucleotide Polymorphisms

FinnGen

Finnish National Genome Project

SE

Standard Error

SBP

Systolic blood pressure

OR

odds ratio

CI

confidence interval