STRENGTHS AND LIMITATIONS OF THIS STUDY
-
Large and diverse sample size.
-
Comprehensive adjustment for multiple potential confounders.
-
Cross-sectional design limits causal inference.
-
Reliance on self-reported obstructive sleep apnoea diagnosis.
-
Single measurement of 25-hydroxyvitamin D may not capture seasonal variations.
Introduction
Obstructive sleep apnoea (OSA) is a prevalent sleep-breathing disorder, diagnosed by an Apnoea-Hypopnea Index (AHI) of ≥5/hour.1 It is estimated that up to 40% of individuals classified as obese suffer from OSA, with obesity being a major risk factor for the disorder.2 Furthermore, the prevalence of OSA is high among those who have undergone bariatric surgery, with studies reporting rates as high as 80.5%.3 OSA is not only prevalent but also poses significant health risks that extend beyond sleep disturbances. OSA is associated with numerous adverse health outcomes, including cardiovascular diseases, metabolic disorders and neurocognitive impairments.4 The intermittent hypoxia and sleep fragmentation characteristic of OSA contribute to systemic inflammation, oxidative stress and sympathetic nervous system activation, which in turn, exacerbate these comorbid conditions. Given its widespread prevalence and profound impact on health, it is crucial to explore modifiable risk factors, such as vitamin D deficiency, that may offer potential avenues for intervention and improved management of OSA.
In addition to its role in bone homeostasis,5 25-hydroxyvitamin D (25(OH)D) may also play a role in regulating sleep. This is because vitamin D receptors are present in brain regions that are involved in sleep regulation, such as the hypothalamus, prefrontal cortex, substantia nigra and nucleus accumbens.6 Furthermore, there is evidence to suggest that 25(OH)D deficiency is associated with metabolic disorders similar to OSA, including impaired glucose metabolism, obesity and bone deformities.7
Despite several epidemiological studies suggesting a link between low vitamin D concentrations and OSA, a definitive relationship between 25(OH)D and OSA has yet to be established, and conflicting evidence exists in the literature. For example, a cross-sectional study of 106 Dublin adults classified as either non-OSA or OSA found that lower levels of 25(OH)D were present in the Caucasian OSA population.8 However, a separate study of 121 male OSA patients failed to establish an association between serum 25(OH)D levels and OSA.9 Unfortunately, these studies are subject to various limitations, such as small sample sizes and inadequate control of covariates such as dietary vitamin D intake, physical activity, seasonality and comorbidities. Additionally, it is unclear whether race/ethnicity and obesity status might have an impact on the relationship between 25(OH)D and OSA.
To address these gaps in current knowledge, the aim of this study is to investigate the potential relationship between 25(OH)D and OSA in a population of individuals aged 16 years and older, using a diverse sample with varying ages and ethnicities.
Methods
Study population
This study drew on data obtained from the National Health and Nutrition Examination Survey (NHANES) database, which is sponsored by the Centres for Disease Control and Prevention and operates on a 2-year survey cycle. Beginning in 1999, NHANES has been conducted without interruption. Each year, around 5000 individuals are interviewed in their homes, and physical examinations and biological sample collection are conducted in mobile units. The NHANES survey targets residents from 15 urban areas across the USA, including individuals of African, Asian and Hispanic descent.10
For our study, we analysed NHANES data collected during the 2007–08 survey cycle due to the availability of questionnaire data on sleep apnoea. To ensure data completeness, we excluded participants with missing 25(OH)D data (n=3199) and those with missing responses to the ‘OSA’ question (n=2049). Ultimately, our study sample included 4901 participants (online supplemental figure S1).
Supplemental material
The NHANES procedure was authorised by the Institutional Review Board of the National Centre for Health Statistics. Informed consent was obtained from all NHANES participants or their legal representatives (<18 years).11
Obstructive sleep apnea
As a component of the NHANES study, sleep apnoea diagnosis was based on medical condition questionnaires collected through interviews. Participants were asked whether a doctor had ever diagnosed them with a sleep disorder. If so, they were asked to specify the type of sleep disorder from among categories such as OSA, insomnia, restless legs syndrome or other (online supplemental materials).
Serum 25(OH)D
The common gauge of vitamin D nutritional status is the 25(OH)D indicator. Blood samples were obtained from each participant at a mobile examination centre and immediately frozen at 30°C to measure serum 25(OH)D concentrations. For the 2007–08 cycles, chromatography-tandem mass spectrometry was used to determine serum 25(OH)D.
Covariates
Study subjects received questionnaires under the guidance of a professional investigator. The following data associations with 25(OH)D and/or OSA were collected by interview questionnaires: age, sex, race, education level, income to poverty ratio, body mass index (BMI),12 smoking status,13 alcohol use,14 diabetes15 and cardiovascular disease.16
Statistical analysis
The β and accompanying 95% CI for the connection between 25(OH)D and OSA were calculated using a multivariable logistic regression model. In accordance with STROBE’s (Strengthening the Reporting of Observational Studies in Epidemiology) criteria,17 we created three models. The first was a basic model that simply included adjustments for age, sex and race (model 1). BMI was further adjusted in model 2 before being entirely adjusted in model 3, which also took into account all other covariables. To investigate potential modifying variables, subgroup analyses with stratification by age, sex, BMI and race were conducted. A p value of 0.05 was used to determine statistical significance for all analyses, which were performed using the Empower Stats programme and R V.3.4.3.
Results
Study sample
Table 1 provides a breakdown of the study sample, comparing those with and without OSA. The OSA group had a mean age of 53.85±13.87 years, which was older than the non-OSA group (44.81±17.97 years). Additionally, the percentage of men in the OSA group (64.53%) was higher than that of women (47.69%). Significant differences (p<0.05) were found between the two groups in terms of 25(OH)D levels, race, education level, income to poverty ratio, BMI, smoking status, alcohol use, diabetes status and cardiovascular disease.
Characteristic of study sample with and without OSA
Multiple regression model
Table 2 displays the results of the regression analyses investigating the association between 25(OH)D and OSA. After adjusting for age, sex and race (model 1), a significant negative association was observed (β=−3.21, 95% CI −6.17 to –0.26). However, when BMI was adjusted as an additional covariate (model 2), this relationship was no longer significant (β=1.47, 95% CI −1.48, 4.42). In the fully adjusted model (model 3), which adjusting all potential confounding variables, there was no significant association between OSA and 25(OH)D (β=0.92, 95% CI −1.93, 3.76).
Associations between OSA and serum vitamin D
Subgroup analyses
When stratified by sex, age and race (tables 3–5), no significant associations between OSA and 25(OH)D were found in the 16–19 years group (β=7.99, 95% CI −21.42, 37.40), 20–59 years group (β=2.34, 95% CI −1.53, 6.20) or 60–85 years group (β=−0.25, 95% CI −4.47, 3.96). Similarly, there were no significant associations observed in either males (β=−0.65, 95% CI −1.51, 0.21) or females (β=0.65, 95% CI −0.76, 2.07) nor in non-Hispanic White (β=1.96, 95% CI −2.10, 6.03), non-Hispanic Black (β=1.52, 95% CI −4.11, 7.16), Mexican American (β=−6.39, 95% CI −13.85, 1.06) or other race (β=3.58, 95% CI −4.69, 11.85) groups. Additionally, there was no significant association observed in participants with BMI <25 (kg/m²) (β=4.48, 95% CI −4.69, 11.85), those with BMI inside 25–29.9 (kg/m²) (β=−0.03, 95% CI −5.49, 5.43) or those with BMI ≥30 (kg/m²) (β=−1.37, 95% CI −4.50, 1.76) (online supplemental table S1). Furthermore, the impact of OSA on 25(OH)D was not found to be affected by sex, age, race or BMI (p for all interactions >0.05).
Associations between OSA and serum vitamin D stratified by age
Associations between OSA and serum vitamin D stratified by sex
Associations between OSA and serum vitamin D stratified by race
Discussion
After adjusting for BMI and other covariates (sex, education level, income to poverty ratio, smoking status, alcohol use, diabetes and cardiovascular disease), the statistically significant association between 25(OH)D and OSA was no longer observed. This non-significance was further confirmed in all subanalyses stratified by age, gender, race and BMI.
Our study results contradict earlier research conducted by Mete et al who observed a lower vitamin D level in OSA patients with an AHI ≥30/hour compared with those with an AHI ≤5/hour.18 However, it should be noted that Mete et al’s study focused solely on obese patients, thus, failing to consider the impact of BMI on the association between 25(OH)D and AHI. In contrast, two recent studies, which controlled for BMI, found no evidence linking 25(OH)D to OSA.9 19 These findings are consistent with our study. After controlling for potential confounding variables such as BMI, the authors of a study that included 121 adults performed multinomial regression analysis and found no significant association between the severity of OSA and vitamin D levels. The results of the correlation analysis also revealed a non-significant relationship between vitamin D levels and Apnoea–Hypopnoea Index (r=0.017, p=0.877). In light of these findings, it can be inferred that vitamin D status does not impact the severity of OSA.9 Similarly, a Mendelian randomisation study determined that there was no significant association between 25(OH)D and OSA after controlling for BMI and other relevant confounding variables.19 These findings suggest that the reduction in 25(OH)D levels in OSA patients may be due to a common confounding factor, BMI. Considering the differences in 25(OH)D levels among various races,20 subgroup analyses were performed. In the stratified analysis of race, 25(OH)D was not found to be associated with OSA among non-Hispanic White, non-Hispanic Black, Mexican American or other races. Similarly, consistent outcomes were evident in the subgroup analyses conducted based on age, gender and BMI.
Obesity represents one of the prevalent risk factors shared by both 25(OH)D deficiency and OSA. Our findings further reveal that BMI significantly impacts the association between 25(OH)D and OSA. Therefore, when investigating the link between 25(OH)D and OSA, BMI needs to be considered, with the following explanations serving as partial justification. One plausible explanation is that obese individuals possess greater adipose tissue mass, leading to increased 25(OH)D uptake and storage.21 Additionally, a volume dilution effect may come into play, whereby 25(OH)D is dispersed across a larger area of adipose tissue, liver and muscle.22 At the same time, obese individuals may frequently inhabit indoor settings, limiting their exposure to sunlight and potentially resulting in decreased synthesis of 25(OH)D.23
To the best of our knowledge, this study represents the most comprehensive investigation of the correlation between OSA and 25(OH)D levels in a multiethnic sample, which can serve as a surrogate for assessing the entire population. Moreover, the extensive data collected in the NHANES allowed us to consider an array of potential confounders, including dietary vitamin D intake, physical activity, seasonality and comorbidities. There are also some limitations worth noting. First, the diagnosis of OSA relied on self-reported symptoms of sleep apnoea, without objective measurements, which may have led to misclassification of OSA. Misclassification due to undiagnosed mild to moderate OSA may lead to underestimation of the association if mild to moderate OSA increases the risk of vitamin D deficiency. However, the study used a large sample from NHANES surveys, resulting in broad coverage and representativeness. Additionally, self-administered questionnaires have been widely used for evaluating self-reported OSA.13 24 The high prevalence of OSA in the sample under study, which closely matches previous population prevalence estimates of sleep apnoea syndrome,25 lends support to the accuracy of health-professional-diagnosed self-report as a diagnostic tool.26 27 Second, a significant factor affecting endogenous vitamin D levels was not accounted for in the research design. Geographical location is an important determinant of sunlight exposure,18 28 but this information was not collected. However, a previous study found no significant association between geographical location and 25(OH)D levels, suggesting that the impact of geographical location on 25(OH)D levels may be negligible.29 It is worth noting that 25(OH)D levels exhibit seasonal variation,30 and a single 25(OH)D measurement may not capture the full spectrum of year-round 25(OH)D fluctuations. Third, it should be noted that the present study’s cross-sectional methodology limits its ability to determine the causal relationship between 25(OH)D and OSA. Fourth, the investigation into the relationship between different degrees of OSA and 25(OH)D was not conducted, as the severity of OSA was not assessed in the present study. Fifth, while 25(OH)D is considered a superior marker for assessing vitamin D status, future research could explore the use of bioavailable vitamin D as an alternative marker.20 31 Overall, these limitations must be taken into account when interpreting the findings of the present study.
Conclusions
Our study did not find a significant association between OSA and 25(OH)D. However, it is possible that the observed association between lower 25(OH)D levels and OSA could be attributed to confounding factors such as higher BMI levels in the OSA group. In order to prevent 25(OH)D insufficiency in individuals with OSA, there is a need to improve the management of obesity. Thus, our study highlights the importance of comprehensive management of both OSA and obesity for promoting optimal health outcomes.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. The survey data are publicly available on the internet for data users and researchers throughout the world (www.cdc.gov/nchs/nhanes/).
Ethics statements
Patient consent for publication
Ethics approval
The ethics review board of the National Center for Health Statistics approved all NHANES protocols and written informed consents were obtained from all participants or their proxies (<18 years). The study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Bioethics Committee of Southern Medical University reviewed our study and have waived the need for ethical approval.
Acknowledgments
Thanks to Jing Zhang (Shanghai Tongren Hospital) for his work on the NHANES database.
His outstanding work, nhanesR package and webpage, makes it easier for us to explore the NHANES database.
This post was originally published on https://bmjopen.bmj.com