Author: admin

Introduction

Behçet’s disease (BD) is a systemic vasculitis frequently associated with intraocular inflammation. It is characterized by significant clinical heterogeneity and presents with various systemic manifestations, including mucocutaneous, articular, vascular, neurological, and gastrointestinal features. It is most common in regions along the historic “Silk Road”, stretching from Eastern Asia to the Mediterranean basin. The highest incidence has been recorded in Turkey.^1,2

The etiopathogenesis of BD has not been clarified. It usually affects young adults 20 to 40 years of age and is characterized by a relapsing and remitting course.² While both genders are affected equally, male patients tend to experience a more severe progression of the disease.

The criteria established by the International Study Group for Behçet’s disease requires that patients must have recurrent oral ulcers along with at least two of the following criteria: recurrent genital ulcers, skin lesions, eye lesions, or a positive pathergy test.³

Ocular involvement contributes most to the morbidity in BD. Common eye symptoms include blurred vision, reduced visual clarity, redness, periocular or periorbital pain, light sensitivity, tearing, foreign body sensation, and headaches. The most common type of ocular involvement in BD is non-granulomatous uveitis, often accompanied by retinal vasculitis and can be the initial manifestation of the disease. Bilateral involvement is generally observed and may affect anterior, posterior or both segments of the eye (panuveitis). Panuveitis and posterior uveitis are the most common forms of involvement, posing a significant threat to vision and a high risk of lasting complications.⁴

In recent years, research in Behçet’s disease uveitis (BU) has evolved significantly, particularly through the integration of artificial intelligence (AI) and radiomics into ophthalmic diagnostics. Novel AI-driven models, especially those based on OCT angiography and radiomic biomarkers, have demonstrated high diagnostic performance in identifying BU, providing a promising complement to clinical evaluation and overcoming.⁵

The primary goals in the management of patients are to achieve rapid resolution of intraocular inflammation, to prevent recurrent attacks, and to ensure complete remission and preservation of vision.

Patients with isolated anterior uveitis can be managed using topical steroids. However, systemic IMT (immunomodulatory therapy), such as azathioprine (AZA) could be considered in cases of poor prognostic factors such as hypopyon, early onset of the disease and male gender.

In case of posterior uveitis and panuveitis systemic IMT should be started as soon as possible, such as azathioprine and cyclosporine, in complement to oral corticosteroids.^3,6

Intra or periocular corticosteroids can be used in addition to systemic treatment for unilateral exacerbations.⁷

In refractory or recurrent cases, biological therapies such as infliximab (IFX) and adalimumab (ADA) are recommended.⁸ However, the treatment strategy involving conventional immunomodulators versus anti-TNF agents has not been clearly defined.⁹

Determining whether disease stability results from the treatment itself or the naturally relapsing-remitting course of the disease is challenging. Also, establishing a standardized protocol for discontinuing treatment in all BD patients is complicated because of the higher relapse risk in men and younger individuals. There is no consensus on the appropriate timing to discontinue treatment for BD patients in remission. Current recommendations suggest that in patients with significant organ involvement, as in the case of uveitis, remission may be achieved after 2–5 years of IMT⁶ or after more than 6 years.¹⁰

This strategy aims to maintain remission and limit ocular inflammation. Treatment efficacy is critical to avoid the consequences of long-term ocular inflammation, associated with significant morbidity among an overall young active population, including permanent vision loss. BD uveitis is responsible for an important share of the world’s immune-related blindness, so initiating treatment as early as possible can improve this prognosis.⁷

Materials and Methods

Study Design, Setting and Participants

This is a retrospective, single center, longitudinal study of patients with unilateral and bilateral uveitis related to Behçet’s disease, followed in the Ophthalmology department of Centro Hospitalar Universitário São João (Porto, Portugal).

Data from all patients under systemic IMT evaluated in the past 10 years were collected by chart review. Initial screening searched for patients under methotrexate, adalimumab, cyclosporine, azathioprine, infliximab and certolizumab, which provided a total of 509 medical processes.

Afterwards, only patients with Behçet’s disease were included.

Data Collection

The following information was extracted from each case, based on the patient’s electronic medical records and procedure reports: demographic data, characterization of the initial uveitis episode, type of uveitis, total follow period, type and duration of IMT, need for adjuvant corticosteroid therapy and pattern of disease remission and relapses were recorded. Intolerance or toxicity as well as treatment’s discontinuation were also documented.

Statistical Analysis

Kolmogorov–Smirnov and Shapiro–wilk tests were used to assess whether each continuous variable followed a normal distribution, with Shapiro–Wilk being preferred for small sample sizes. Normally distributed data is reported as mean ± standard deviation (SD) while non-normally distributed data is reported as median and interquartile range (IQR). Categorical variables are presented as absolute number and percentage. Parametric tests like student’s t-test and non-parametric tests like Mann–Whitney or Wilcoxon were used for variables comparison between groups, according to the normality of data. We used Mann–Whitney to compare independent variables and Wilcoxon to compare paired (dependent) variables. Categorical variables were compared using Chi-square or Fishers exact tests.

A p-value<0.05 was considered statistically significant. Statistical analysis was done using IBM SPSS^® software (version 26.0).

Results

We analyzed 38 patients with BD under IMT.

Demographic Features and Uveitis Characterization

The mean total follow-up time of patients in this sample was 122.5 ± 62.6 months [10–250]. Patients’ mean age was 44.1 ± 11.6 [24–70] and 20 patients were women (52.6%).

Fourteen (36.8%) presented with anterior uveitis, which include iritis, iridocyclitis and anterior cyclitis. Three patients (7.9%) were diagnosed with posterior uveitis, which included choroiditis and/or retinitis. While retinal vasculitis could technically be classified as posterior uveitis, we opted to distinguish it as a separate entity since twelve patients (31.6%) presented specifically with vasculitis. Three patients exhibited with intermediate uveitis (7.9%), which included pars planitis, posterior cyclitis and hyalitis, and five patients had panuveitis (13.2%), which included anterior chamber, vitreous and retina or choroid. One presented with optic neuritis (2.6%).^11,12 Twenty-four patients (63.2%) presented with bilateral uveitis and twenty-six patients (68.4%) had uveitis as their first presentation of the disease (Table 1).

Treatment

The median duration of treatment was 63.50 ± 59.1 [4–201] months. Among the patients, 24 (63.2%) received IMT for at least 48 months, and 16 (50%) were treated for a minimum of 72 months.

Azathioprine was the most widely used immunomodulatory agent (n = 14, 36.8%). Cyclosporine was the second mostly used (n = 11, 28.9%). Adalimumab was used in 4 patients, (10.5%) of which three were already on cyclosporine and one on azathioprine and infliximab. Infliximab was used in 3 patients (7.9%), of which one was already on cyclosporine, and methotrexate was used in 2 patients (5.3%). Ocular inflammation was effectively managed with a combination of methotrexate and adalimumab in 2 patients (5.3%). In one patient (2.6%), inflammation was controlled using a combination of infliximab and azathioprine. Additionally, azathioprine and cyclosporine were administered in 1 patient (2.6%). Twenty-two patients (57.9%) effectively managed ocular inflammation after the first immunomodulator.

Before IMT patients presented with a median of 2 ± 2.0 [0–10] relapses per year, a number that significantly decreased to a median of 1 ±1.2 [0–4] with the introduction of IMT (p< 0.001).

Twenty-eight patients only used non-biological treatment (methotrexate, cyclosporine and azathioprine), 7 patients only used biological treatment (infliximab and adalimumab) and 3 patients used a combination between biological and non-biological treatment. There was a slight superiority of biological IMT compared to non-biological IMT in reducing the number of recurrences after treatment (p = 0.045).

Seventeen patients with only non-biological treatment needed adjuvant oral corticosteroids and no patients with biological treatment needed adjuvant oral corticosteroids. Two patients (5.3%) used periocular corticosteroids (Table 2).

Table 2 Treatment and Inflammation Control. Comparison Between the Number of Relapses per year Before and After IMT

In Table 3 LogMAR (Logarithm of the minimum angle of resolution) is the scale used to quantify visual acuity in this population. There was no correlation between LogMAR and the severity of uveitis in Behçet’s disease in our sample.

Table 3 Visual Acuity After Starting Systemic Immunomodulatory Therapy- LogMAR

Treatment Discontinuation and Relapse Profile

Sixteen patients (42.1%) stopped treatment: 6 cases (37.5%) as medical based decision because of long-term remission of the disease, 5 cases (31.3%) because of loss of follow-up, 3 cases (18.8%) because of side effects of the medication and 2 cases (12.5%) because of patient unadvised decision.

Patients who discontinued treatment based on medical decisions had a median treatment duration of 77 months. Those who stopped due to side effects were treated for a median of 43 months. Similarly, patients who discontinued treatment because of loss of follow-up or unadvised decision had a median treatment duration of 48 months (Table 4).

Table 4 Incidence of Treatment Discontinuation and Relapse Profile

Patients were monitored for a median of 57 ± 46.4 months after stopping treatment, during which 4 (30.8%) out of 16 patients developed recurrence after a median period of 13 ± 10.4 months, range [2–27]. Two of these patients were treated for less than 4 years and all of them were treated for less than 6 years.

Eight of the 12 patients who did not relapse had been treated for more than 4 years, and 4 of the 12 patients who did not relapse had been treated for more than 6 years, without any predominance regarding the type of IMT used. Five of the 6 patients who stopped IMT because of a medical based decision had been treated for a minimum of 4 years and no relapse of ocular inflammation occurred among them.

Discussion

This study aims to provide a comprehensive review of the most significant aspects of systemic IMT for uveitis in the context of Behçet’s disease. Behçet’s disease uveitis may frequently lead to blindness when left uncontrolled or inadequately treated. Early initiation of appropriate treatment is crucial to improving the prognosis and preventing vision loss and ocular complications.¹³ Clear therapeutic guidelines and protocols regarding the appropriate duration of IMT for patients with BD uveitis have not yet been established.

Our study presented a mean follow-up period of patients with Behçet’s uveitis of around 10 years. This extensive follow-up period provides valuable insight into the long-term effectiveness of appropriate treatment in controlling the disease. It allows us to see the potential of preventing recurring episodes of intraocular inflammation, which could result in severe complications and permanent visual impairment.

The mean age of the patients was 44 years, with a range between 24 and 70 years, which reflects individuals with Bechet’s disease who are monitored at the hospital over varying follow-up periods, highlighting the broad age distribution within this population. The gender distribution in this population was balanced, with 47.4% men and 52.6% women.

Regarding ocular manifestations, anterior uveitis was the most common isolated presentation. Several factors, including, early detection of ocular involvement, and possible regional variations in clinical phenotype may influence this finding.

As a hallmark of Behçet’s disease, vasculitis was the second most common type of inflammatory ocular disease. Also, if posterior uveitis is split into its usual subcategories—choroiditis, retinitis, and vasculitis—the combined cases of posterior segment involvement exceed those of anterior uveitis. Therefore, it is of note the predominance of posterior segment involvement, often in the form of retinal vasculitis, which is in line with other publications.¹ Panuveitis is the third most common type of inflammatory ocular disease. Additionally, there is a high proportion of bilateral involvement, which aligns with findings from other studies.¹

Also, in two-thirds of cases uveitis was the first manifestation of the disease, emphasizing the potential for the screening of early ocular signs as an important diagnostic indicator in Behçet’s disease. An early diagnosis often correlates with a more favorable prognosis.

IMT significantly reduces the frequency of relapses as a reflex of its improved control of disease activity in Behçet’s disease, preventing long-term complications, including organ damage and disability, which is consistent with existing literature. Azathioprine was the most used systemic immunomodulatory drug for the treatment of uveitis in BD and cyclosporine was the second most widely used agent. Several studies agree with azathioprine and cyclosporine as first-line immunosuppressive options for uveitis in Behçet’s disease.^14,15 The introduction of biologic therapies such as infliximab and adalimumab after the failure of non-biological immunomodulatory agents (azathioprine and cyclosporine), are deemed a better approach for refractory cases.^8,16 The combination of biological and non-biological therapies, such as methotrexate and adalimumab or infliximab and azathioprine can be used in certain cases of treatment-resistant uveitis with the intention of reducing the immunogenicity of these biological agents and thus improving their efficacy. More than half the patients were controlled after the first immunomodulator. There was a slight, but not clear, superiority of biological IMT compared to non-biological IMT in reducing the number of recurrences after treatment.¹⁷ This comparison should be interpreted with caution due to the small sample of the biological subgroup.

Periocular corticosteroids injections were needed in 2 patients as adjunctive therapy because of acute severe recurrences to alleviate symptoms. Patients on non-biological treatments only required more adjuvant corticosteroids in comparison to those on biological therapies. This finding suggests that biological treatment may act as superior corticosteroid-sparing agents, effectively controlling ocular inflammation without the need for prolonged corticosteroid use, and usually with better tolerance (fewer well-documented side effects).^16,18

Regarding tolerance, only 3 of 38 patients stopped treatment. One patient discontinued treatment because of side effects from azathioprine, including diarrhea and hepatic toxicity, while two others chose to stop treatment due to a desire to become pregnant. This represents an overall favorable safety profile.

In our study, about two thirds were treated for more than 4 years and half were treated for more than 6 years. IMT was associated with a statistically significant decrease in the number of relapses per year, defined as an increase in inflammatory activity following a period of remission.

The patients who stopped IMT because of medical decision (6 out of 38 patients) had a median treatment duration around 6 years. Eight out of twelve patients who stopped treatment and were treated for more than 4 years did not relapse and all the patients who stopped treatment and were treated for more than 6 years (4 patients) did not relapse.

Regarding the 4 patients (30.8%) with relapse after stopping treatment, 2 of these have been treated for less than 4 years and all of them had been treated for less than 6 years. The median time for recurrence after discontinuation was 13 months, ranging from 2–27 months, and it did not appear to be influenced by the duration of treatment among those who relapsed (maximum 66 months). This result must be outlined in the median period of surveillance of 57 months after IMT cessation.

The medical decision to stop IMT after 4–6 years of complete remission appears to be a safe approach, especially in patients who were treated for 6 years, without increasing the incidence of recurrence.

Our study limitations include its small sample size and single-center design. The data collected focused on ophthalmological parameters, and information on extraocular manifestations, such as mucocutaneous or visceral lesions, was not consistently or systematically available. This limitation restricts the analysis of the systemic complexity of the disease and should be taken into account when interpreting the results.

Nonetheless, as a tertiary hospital, we were able to analyze a cohort of patients with uveitis in the context of a rare but sight-threatening disease with a significant length of follow-up, which is an advantage regarding long-term evaluation and follow-up of these patients. This kind of revision may possibly shed some light into specific characteristics of each group, thus allowing for the definition of management protocols in the future.

Conclusion

Most of the patients with Behcet’s uveitis needed IMT to effectively control the ocular inflammation and thus achieve durable remission. Azathioprine and cyclosporine were the most used systemic immunomodulatory drugs for the treatment of uveitis in the context of Behcet’s disease and are a safe first line approach for Behcet’s uveitis.

A medical decision to discontinue treatment after 4 to 6 years of sustained inflammation control appears to be safe, particularly in patients who were treated for 6 years.

Abbreviations

BD, behçet disease; IMT, immunomodulatory therapy.

Data Sharing Statement

Access to any information such as the study protocol or anonymized data can be available upon reasonable request.

Ethics/Ethical Approval

The study was approved by the Institutional Ethics Review Board of Centro Hospitalar Universitário de São João, Porto, Portugal. The protocol conformed with the canons of the declaration of Helsinki for research involving human participants, as well the European Union’s General Data Protection Regulation. Informed consent was waived due to the retrospective nature of the study and the protection of patient data confidentiality. This article was redacted according to the recommendations of the Reporting of Studies Conducted using Observational Routinely-collected health Data (RECORD) statement.

Acknowledgments

Only the named authors have collaborated in the writing of this paper.

Author Contributions

All authors contributed to the study conception and design. Material preparation was performed by Luís Figueira, Joana Rodrigues Araújo and Ana Margarida Ferreira. Data collection was performed by Ana Margarida Ferreira and Mariana Almeida. Analysis was performed by Mariana Almeida and Luís Figueira. The first draft of the manuscript was written by Mariana Almeida, and all the authors took part in revising or critically reviewing the article. All authors read and approved the final manuscript.

Funding

The authors declare that they have no financial ties to declare. No funding or sponsors were undertaken in the preparation of the manuscript.

Disclosure

The authors have no conflicts of interest to declare for this work.

References

1. Emmi G, Bettiol A, Hatemi G, Prisco D. Behçet’s syndrome. Lancet. 2024;403(10431):1093–1108. doi:10.1016/S0140-6736(23)02629-6

2. van der Houwen TB, van Hagen PM, van Laar JAM. Immunopathogenesis of Behçet’s disease and treatment modalities. Semin Arthritis Rheum. 2022;52:151956. doi:10.1016/j.semarthrit.2022.151956

3. Tugal-Tutkun I. Behcet’s Uveitis. Middle East Afr J Ophthalmol. 2009;16(4):219–224. doi:10.4103/0974-9233.58425

4. Zhong Z, Su G, Yang P. Risk factors, clinical features and treatment of Behçet’s disease uveitis. Prog Retin Eye Res. 2023;97:101216. doi:10.1016/j.preteyeres.2023.101216

5. Lu A, Li K, Zhang X, Su G, Yang P. Development and validation of novel retina biomarkers and artificial intelligence models for Behçet disease uveitis prediction. Biomed Signal Process Control. 2024;94:106271.

6. Karadag O, Bolek EC. Management of Behcet’s syndrome. Rheumatology. 2020;59(Suppl 3):iii108–iii117. doi:10.1093/rheumatology/keaa086

7. Joubert M, Desbois AC, Domont F, et al. Behçet’s disease uveitis. J Clin Med. 2023;12(11):3648. doi:10.3390/jcm12113648

8. McNally TW, Damato EM, Murray PI, Denniston AK, Barry RJ. An update on the use of biologic therapies in the management of uveitis in Behçet’s disease: a comprehensive review. Orphanet J Rare Dis. 2017;12(1):130. doi:10.1186/s13023-017-0681-6

9. Leclercq M, Langlois V, Girszyn N, et al. Comparison of conventional immunosuppressive drugs versus anti-TNF-α agents in non-infectious non-anterior uveitis. J Autoimmun. 2020;113:102481. doi:10.1016/j.jaut.2020.102481

10. Malek Mahdavi A, Khabbazi A, Hajialilo M. Long-term outcome and predictors of remission in Behçet’s disease in daily practice. Mod Rheumatol. 2021;31(6):1148–1157. doi:10.1080/14397595.2021.1886623

11. Deuter CM, Kötter I, Wallace GR, Murray PI, Stübiger N, Zierhut M. Behçet’s disease: ocular effects and treatment. Prog Retin Eye Res. 2008;27(1):111–136. doi:10.1016/j.preteyeres.2007.09.002

12. Jabs DA, Dick AD, Dunn JP. Classification criteria for Behçet disease uveitis. Am J Ophthalmol. 2021;228:80–88. doi:10.1016/j.ajo.2021.03.058

13. Dick AD, Tundia N, Sorg R, et al. Risk of ocular complications in patients with noninfectious intermediate uveitis, posterior uveitis, or panuveitis. Ophthalmology. 2016;123(3):655–662. doi:10.1016/j.ophtha.2015.10.028

14. Alibaz-Oner F, Direskeneli H. Advances in the treatment of Behcet’s disease. Curr Rheumatol Rep. 2021;23(6):47. doi:10.1007/s11926-021-01011-z

15. Hatemi G, Christensen R, Bang D, et al. 2018 update of the EULAR recommendations for the management of Behçet’s syndrome. Ann Rheum Dis. 2018;77(6):808–818. doi:10.1136/annrheumdis-2018-213225

16. Vallet H, Riviere S, Sanna A, et al. Efficacy of anti-TNF alpha in severe and/or refractory Behçet’s disease: multicenter study of 124 patients. J Autoimmun. 2015;62:67–74. doi:10.1016/j.jaut.2015.06.005

17. Urruticoechea-Arana A, Cobo-Ibáñez T, Villaverde-García V, et al. Efficacy and safety of biological therapy compared to synthetic immunomodulatory drugs or placebo in the treatment of Behçet’s disease associated uveitis: a systematic review. Rheumatol Int. 2019;39(1):47–58. doi:10.1007/s00296-018-4193-z

18. Li B, Li H, Huang Q, Zheng Y. Optimizing glucocorticoid therapy for Behçet’s uveitis: efficacy, adverse effects, and advances in combination approaches. Int Ophthalmol. 2023;43(11):4373–4381. doi:10.1007/s10792-023-02808-w

Gut bacteria are fundamental to many aspects of human health, from digestion to immunity. Yet, the complexities of these microbial communities – along with the vast number of different species and metabolites they produce – make it challenging to study how they interact with the body.

Researchers at the University of Tokyo have applied a type of artificial intelligence (AI) technique to explore large datasets on gut bacteria, aiming to uncover relationships that traditional analytical tools have struggled to reveal.

Their advance is published in Briefings in Bioinformatics.

Challenges in mapping microbial interactions

The human body hosts around 30 to 40 trillion cells, yet the intestines house approximately 100 trillion bacteria. These microbes play critical roles in digesting food, but they also influence metabolism, immune responses, and even mental health.

The bacteria produce a wide variety of metabolites, which act as molecular messengers throughout the body. Understanding the intricate relationships between these bacteria and their metabolites could open the door to personalized treatments for a range of health conditions.

“The problem is that we’re only beginning to understand which bacteria produce which human metabolites and how these relationships change in different diseases,” said project researcher Tung Dang, from the Tsunoda lab in the Department of Biological Sciences. “By accurately mapping these bacteria-chemical relationships, we could potentially develop personalized treatments. Imagine being able to grow a specific bacterium to produce beneficial human metabolites or designing targeted therapies that modify these metabolites to treat diseases.”

However, identifying meaningful patterns within the vast amounts of data generated by gut microbiome studies is a complex task. The sheer number of bacteria and metabolites involved, combined with their interactions, presents a formidable analytical problem.

Bayesian neural networks to the rescue

To tackle this challenge, Dang and his team began to explore whether state-of-the-art AI tools could be applied to this problem. The result – a variable Bayesian neural network model known as VBayesMM.

“Our system, VBayesMM, automatically distinguishes the key players that significantly influence metabolites from the vast background of less relevant microbes, while also acknowledging uncertainty about the predicted relationships, rather than providing overconfident but potentially wrong answers,” said Dang.

“When tested on real data from sleep disorder, obesity and cancer studies, our approach consistently outperformed existing methods and identified specific bacterial families that align with known biological processes, giving confidence that it discovers real biological relationships rather than meaningless statistical patterns,” said Dang.

One of the main advantages of VBayesMM is that it can handle and communicate uncertainty, which can give researchers more confidence in its outputs than a tool which cannot. The system is also optimized to cope with heavy analytical workloads, such as the huge datasets that must be processed to understand the gut microbiome.

Limitations and future improvements

Despite its advantages, the system is not without limitations. One key challenge is the need for more detailed bacterial data compared to the metabolites they produce. When the available bacterial data is insufficient, the system’s accuracy drops. Additionally, the model assumes that microbes act independently, though in reality, they interact in highly complex ways, making it difficult to model these relationships fully. Despite its optimization for heavy workloads, the system does still carry a relatively high computational cost which may be a barrier to some groups.

Looking ahead, Dang and his team plan to integrate more comprehensive datasets that include a broader range of bacterial metabolites.

“We plan to work with more comprehensive chemical datasets that capture the complete range of bacterial products, though this creates new challenges in determining whether chemicals come from bacteria, the human body or external sources like diet,” said Dang.

“We also aim to make VBayesMM more robust when analyzing diverse patient populations, incorporating bacterial ‘family tree’ relationships to make better predictions, and further reducing the computational time needed for analysis.”

With these improvements and adjustments, the team hopes that the insights gained from this work could lead to new clinical treatments based on the manipulation of the microbiome.

“For clinical applications, the ultimate goal is identifying specific bacterial targets for treatments or dietary interventions that could actually help patients, moving from basic research toward practical medical applications,” Dang said.

Reference: Dang T, Lysenko A, Boroevich KA, Tsunoda T. VBayesMM: variational Bayesian neural network to prioritize important relationships of high-dimensional microbiome multiomics data. Brief Bioinform. 2025;26(4). doi: 10.1093/bib/bbaf300

This article has been republished from the following materials. Note: material may have been edited for length and content. For further information, please contact the cited source. Our press release publishing policy can be accessed here.

This content includes text that has been generated with the assistance of AI. Technology Networks’ AI policy can be found here.

Background

Obstructive sleep apnea (OSA), characterized by recurrent upper airway collapse during sleep, leads to chronic intermittent hypoxia (CIH), sleep fragmentation, and systemic pathophysiological changes.^1–3 Globally affecting nearly 1 billion people,⁴ OSA poses a growing public health challenge.⁵

In recent years, the bidirectional link between OSA and endocrine/metabolic disorders has gained significant attention,⁶ particularly the nuanced relationship between OSA and thyroid function.⁷ Thyroid hormones, regulated by the hypothalamic-pituitary-thyroid (HPT) axis, play critical roles in metabolism and cardiovascular function.^8–11 Notably, the evidence regarding the association between OSA and thyroid function remains inconsistent. Some studies suggest a subtle connection, such as hypothyroidism exacerbating OSA through airway narrowing⁷ and OSA-induced oxidative stress impairing thyroid hormone synthesis,^12–15 other research, however, indicates no significant associations.¹⁶ Our previous study revealed that the progression of OSA may promote increased levels of TH, especially FT3, in non-elderly individuals.¹⁷ Adding to this inconsistency, several studies have reported an association between hypothyroidism and OSA severity.¹⁸ Together, these divergent findings highlight the need to clarify the contentious relationship between OSA and thyroid function, including its specific mechanisms and bidirectional interactions.

The sleep apnea-specific hypoxic burden (SASHB) and sleep breathing impairment index (SBII) are crucial early indicators for assessing hypoxia in OSA, comprehensively capturing respiratory event frequency, depth, and hypoxia duration.^19,20 They can reflect the physiological impact of CIH more accurately than the traditional AHI.²¹ This focus on oxygen saturation dynamics is particularly relevant for studying thyroid function, as thyroid hormone synthesis and release are highly sensitive to hypoxic stress. By integrating hypoxia intensity and duration, SASHB and SBII offer a more precise tool for dissecting how CIH modulates thyroid hormone levels. Notably, SASHB has already been linked to glucose and lipid metabolism abnormalities²² and cardiovascular risks,^23,24 underscoring its utility in capturing hypoxia-related endocrine and metabolic dysfunction. However, no relevant studies have been conducted on the relationships among the SASHB, the SBII, and thyroid function.

Sleep consists of rapid eye movement (REM) and non-rapid eye movement (NREM) stages,²⁵ with distinct physiological profiles influencing respiratory function. REM sleep, characterized by heightened brain activity and muscle atonia,^25–27 contrasts with NREM sleep, where physiological functions like heart rate and respiration slow, promoting recovery.^25,26 Patients with OSA exhibit more severe airway collapse during REM sleep, likely due to stage-specific neuromuscular regulation.²⁸ Therefore, it is crucial to consider the impact of different sleep stages on physiological functions when studying the relationship between OSA and thyroid function.

Therefore, this study aimed to explore the thyroid function changes in patients with OSA and analyze the intrinsic connections between SASHB, SBII, and thyroid function indicators (such as thyroid hormones and antibodies). Additionally, this study focused on the associations between the SASHB and SBII during different sleep stages and thyroid function, which have never been explored previously. This research is of great theoretical and practical importance for comprehensively understanding the mechanisms by which OSA affects thyroid function, optimizing clinical diagnostic strategies, and developing targeted therapeutic interventions.

Subjects and Methods

Study Design and Participants

This retrospective study included 1681 individuals with suspected OSA who visited the Department of Otorhinolaryngology at Xi’an Jiaotong University Second Affiliated Hospital from August 2017 to March 2024. Participants were included if they (1) were ≥18 years of age; (2) had undergone overnight polysomnography (PSG) and were diagnosed with OSA; and (3) had complete TH and antibody test results. The exclusion criteria were as follows: (1) history of OSA treatment; (2) severe systemic diseases, such as heart, liver, lung, and kidney failure; (3) other non-OSA sleep disorders; (4) severe mental illnesses or malignant tumors; (5) use of sedatives or medications that may interfere with thyroid function; (6) missing clinical PSG data. After stringent screening, 452 participants with complete data were included in this study. The entire recruitment process is illustrated in Figure 1.

This study strictly adhered to the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of Xi’an Jiaotong University Second Affiliated Hospital (Approval No. 2022–1417), with all participants providing informed consent.

Data Elements

A total of 35 relevant clinical parameters were collected in this study, including the following candidate variables: (1) demographic characteristics, including sex and age; (2) anthropometric measures, including body mass index (BMI), neck circumference (NC) and waist circumference (WC), (3) comorbidities, including history of diabetes, coronary heart disease (CHD) and hypertension; (4) lifestyle habits, including smoking and alcohol use; (5) OSA-related history and indicators, including total sleep time (TST) recorded during the overnight PSG study, Epworth Sleepiness Scale (ESS),²⁹ mean apnea duration, maximum apnea duration, apnea–hypopnea index (AHI), lowest nocturnal peripheral oxygen saturation (Lowest SpO₂), time spent with peripheral oxygen saturation <90% (T90), percentage of time spent with peripheral oxygen saturation <90% (CT90), average heart rate during sleep, lowest heart rate during sleep, highest heart rate during sleep, SASHB, SASHB during NREM sleep (NREM-SASHB), SASHB during REM sleep (REM-SASHB); SBII, SBII during NREM sleep (NREM-SBII), and SBII during REM sleep (REM-SBII); (6) thyroid function-related indicators, including serum free triiodothyronine (FT3, pmol/L), serum free thyroxine (FT4, pmol/L), serum total triiodothyronine (TT3, nmol/L), serum total thyroxine (TT4, nmol/L), serum thyroid stimulating hormone (TSH, mIU/L), thyroid peroxidase antibodies (Anti-TPO, IU/mL), thyroid globulin antibodies (Anti-TG, IU/mL), and reverse triiodothyronine (RT3, ng/dL). All fasting venous blood samples were collected between 7:00 and 8:00 AM on the morning immediately following an overnight PSG study, under strict quality control protocols, with thyroid function indicators analyzed using standardized laboratory procedures.

Sleep Evaluation

To obtain accurate and objective sleep parameters, all enrolled patients underwent overnight PSG monitoring at the Sleep Center of the Department of Otorhinolaryngology-Head and Neck Surgery at the Second Affiliated Hospital of Xi’an Jiaotong University. All the records were evaluated by certified clinical PSG experts, who comprehensively analyzed various parameters, including electroencephalography, electrooculography, electromyography, electrocardiography, nasal and oral airflow recordings, oxygen saturation levels, chest movements, and muscle activity.

The Epworth Sleepiness Scale (ESS) used in this study has been authorized by the copyright holder.

Calculation and Definition

The AHI is defined as the number of apnea and hypopnea events per hour during sleep. In addition, hypopnea is defined as an abnormal respiratory event lasting at least 10s with at least a 30% reduction in thoracoabdominal movement or airflow as compared with baseline and with at least a 4% oxygen desaturation.³⁰ The SASHB is the total area under the baseline SpO₂ curve corresponding to respiratory events per hour during sleep. REM-SASHB and NREM-SASHB refer to the SASHB during REM and NREM sleep, respectively. The SBII is defined as the sum of the duration of breathing events and the corresponding desaturation area per hour during sleep. Accordingly, REM-SBII and NREM-SBII are the SBII during REM and NREM sleep, respectively. In this study, the calculations for SASHB and SBII were based on laboratory test data, including nasal airflow and blood oxygen saturation trend graphs. Using the algorithms developed by Ali Azarbarzin¹⁹ and Wenhao Cao,²⁰ we created calculation codes for SASHB and SBII using Python software (version 3.7), enabling efficient batch processing of the raw data. In this study, when the severity of OSA was evaluated by the AHI, patients were categorized according to the number of events per hour: mild OSA was defined as an AHI of 5 to less than 15, moderate OSA was defined as an AHI of 15 to less than 30, and severe OSA was defined as an AHI of 30 or higher. In addition, when the severity of OSA was assessed by the SASHB and SBII, patients were grouped according to quartiles.

Statistical Analysis

The statistical analyses were conducted using R software (version 4.3.2) and SPSS 26.0 (IBM Corporation, Armonk, NY, USA). If the data were normally distributed, the continuous variables were expressed as the means ± standard deviations; if the data were nonnormally distributed, they were expressed as the medians and interquartile ranges. Categorical variables are presented as counts (n) and percentages (%). First, OSA severity was assessed according to the AHI, SASHB, and SBII, and descriptive statistics were calculated. The Kruskal‒Wallis H-test was used for continuous variables, and the chi-square test was employed for categorical variables to examine intergroup differences in descriptive statistics. Additionally, Spearman correlation analysis was performed to assess the associations between thyroid function parameters and sleep parameters, with Benjamini–Hochberg false discovery rate (FDR) correction applied to account for multiple testing. Furthermore, multiple linear regression analysis was performed to evaluate the independent relationships between sleep parameters and thyroid hormone levels, adjusting for potential confounding factors. Sex-stratified analyses were also performed, with separate multiple linear regression models constructed for male and female subgroups to explore sex-specific associations. Moreover, collinearity diagnostics were conducted before statistical analysis to eliminate potential multicollinearity among the variables. All the statistical tests were two-tailed, with the significance level set at p < 0.05.

Results

Baseline and Sleep Parameter Characteristics

A total of 452 patients with OSA, 395 males and 57 females, were included in this study. Significant baseline differences were observed across AHI severity groups (all p < 0.05, Table 1). The male proportion increased with AHI severity, while BMI, NC, and WC showed an upward trend. Hypertension history, smoking, and alcohol use were more prevalent in the severe OSA group (23.51%, 70.82%, and 81.87%, respectively). Sleep parameters, including ESS, hypoxia duration (T90, CT90), and mean heart rate, worsened with increasing AHI, whereas Lowest SpO₂ decreased progressively.

Table 1 Demographic Characteristics and Sleep Parameters of Patients According to the AHI

Thyroid Function Indicators

The relationship between OSA and thyroid function was explored in depth. When OSA severity was evaluated by the AHI, the results indicated significant differences in FT3, FT4, and TT3 levels among the mild, moderate, and severe OSA groups (all p < 0.05). Specifically, Dunn’s multiple comparison tests revealed significant differences in FT3 between the mild and moderate OSA groups (p < 0.01) and between the mild and severe OSA groups (p < 0.001); FT4 also showed significant differences between the mild and moderate OSA groups (p < 0.01) and between the mild and severe OSA groups (p < 0.01). TT3 differed only between the mild and severe OSA groups (p < 0.05). However, there were no significant differences in TT4, TSH, Anti-TPO, Anti-TG, or RT3 levels (all p > 0.05) (Figure 2 and Table 2).

Table 2 Analysis of Thyroid Indicators According to AHI Severity

Figure 2 Mean values of FT3, FT4, and TT3 at the AHI level. (a) FT3; (b) FT4; (c) TT3. *P < 0.05, **P < 0.01, ****P < 0.0001. Abbreviations: FT3, serum free triiodothyronine; FT4, serum free thyroxine; TT3, serum total triiodothyronine; AHI, apnea–hypopnea index.

In addition to being grouped by the AHI, patients were also classified according to SASHB (≤21.38, 21.38–74.47, 74.47–221.21, and >221.21) and SBII (≤9.53, 9.53–36.05, 36.05–110.34, and >110.34) quartiles. In the analysis of intergroup differences, thyroid function indicators exhibited complex and diverse trends. There were statistically significant differences in the FT3 levels in the SASHB and SBII quartiles (all p < 0.01). With the gradual increase in the SASHB and SBII quartiles, the FT3 levels tended to increase. Although FT4 exhibited a meaningful intergroup difference when grouped by SBII severity (p < 0.05), post hoc multiple comparisons revealed no significant differences in the FT4 levels across groups. Additionally, as the SBII increased, the TSH tended to decrease initially and then increase (Figures 3 and 4, Table 3 and Table 4).

Table 3 Analysis of Thyroid Indicators According to SASHB Severity

Table 4 Analysis of Thyroid Indicators According to SBII Severity

Figure 3 Mean values of FT3 and FT4 at the SASHB level. (a) FT3; (b) FT4. *P < 0.05, **P < 0.01. Abbreviations: FT3, serum free triiodothyronine; FT4, serum free thyroxine; SASHB, sleep apnea-specific hypoxic burden.

Figure 4 Mean values of FT3, FT4, and TSH at the SBII level. (a) FT3; (b) FT4; (c) TSH. *P < 0.05, **P < 0.01. Abbreviations: FT3, serum free triiodothyronine; FT4, serum free thyroxine; TSH, serum thyroid stimulating hormone; SBII, sleep breathing impairment index.

To further explore correlations among the SASHB, the SBII, and thyroid function during the NREM and REM periods, patients were regrouped according to REM-SASHB, NREM-SASHB, REM-SBII, and NREM-SBII severity. Focusing first on NREM sleep, significant differences in FT3, FT4, and TSH levels were found among the groups according to NREM-SASHB and NREM-SBII severity (all p < 0.05), with TSH levels initially decreasing but then increasing. The specific results of the multiple comparison tests are presented in Figure S1. In REM sleep, only FT3 levels differed between the REM-SASHB and REM-SBII groups (all p < 0.05) (Tables S1–S4; Figures S1 and S2).

Exploration of the Correlation of Variables and Regression Analysis

Based on the correlation analysis results after FDR correction, we initially explored the correlations between PSG and thyroid function variables, demonstrating the necessity of regression analysis. FT3 was significantly positively correlated with AHI, mean apnea duration, T90, CT90, REM-SASHB, NREM-SASHB, SASHB, REM-SBII, NREM-SBII and SBII (all q < 0.05), and negatively correlated with lowest SpO₂ (q < 0.05). FT4 was significantly positively correlated with TST and maximum heart rate (both q < 0.05). TT3 was significantly positively correlated with AHI, mean heart rate and maximum heart rate (all q < 0.05). Anti-TPO was significantly positively correlated with TST and T90, and negatively correlated with lowest SpO₂ (all q < 0.05). The negative correlation between TSH and NREM-SBII did not remain significant after correction (q > 0.05), while TT4, Anti-TG and RT3 showed no significant correlations (all q > 0.05). These results further demonstrate the necessity of regression analysis to explore the independent associations (Tables S5 and S6, Figure 5).

Figure 5 Heatmap of the correlations between sleep parameters and thyroid parameters. Abbreviations: TST, total sleep time; AHI, apnea–hypopnea index; Lowest SpO2, lowest oxygen saturation at night; T90, sleep time spent with oxygen saturation below 90%; CT90, the percentage of sleep time with oxygen saturation below 90%; REM-SASHB, sleep apnea-specific hypoxic burden during rapid eye movement sleep; NREM-SASHB, sleep apnea-specific hypoxic burden during non-rapid eye movement sleep; SASHB, sleep apnea-specific hypoxic burden; REM-SBII, sleep breathing impairment index during rapid eye movement sleep; NREM-SBII, sleep breathing impairment index during non-rapid eye movement sleep; SBII, sleep breathing impairment index; FT3, serum free triiodothyronine; FT4, serum free thyroxine; TT3, serum total triiodothyronine; TT4, secretes total thyroxine; TSH, serum thyroid stimulating hormone; Anti-TPO, thyroid peroxidase antibodies; Anti-TG, thyroid globulin antibodies; RT3, reverse triiodothyronine.

After fully adjusting for confounding factors, including age, sex and BMI, SASHB (β = 0.145; p < 0.01), NREM-SASHB (β = 0.127; p < 0.05), REM-SASHB (β = 0.137; p < 0.01), SBII (β = 0.132; p < 0.01), NREM-SBII (β = 0.095; p < 0.05), and REM-SBII (β = 0.145; p < 0.01) were independently associated with elevated FT3 levels, whereas the AHI was not independently associated with FT3 levels (Table 5).

Table 5 Stepwise Multiple Linear Regression for Thyroid Indicators

In the male subgroup, multiple linear regression models were used to evaluate the associations between the AHI, SASHB, SBII, and thyroid function. In the same models, independent correlations were observed between SASHB (β = 0.152; p < 0.01), NREM-SASHB (β = 0.130; p < 0.05), REM-SASHB (β = 0.159; p < 0.01), SBII (β = 0.143; p < 0.01), NREM-SBII (β = 0.103; p < 0.05), and REM-SBII (β = 0.169; p < 0.01) with elevated FT3 levels. However, no significant independent associations were observed between the AHI, SASHB, SBII, and thyroid function indicators in the female subgroup (Tables 6 and 7).

Table 6 Stepwise Multiple Linear Regression Against Thyroid Indicators for Male Subgroups

Table 7 Stepwise Multiple Linear Regression Against Thyroid Indicators for Female Subgroups

Discussion

This study focused on the characteristics of thyroid function in patients with OSA. By analyzing clinical data and PSG results from 452 patients with OSA, we specifically investigated the intrinsic relationships between the SASHB, the SBII, and thyroid function indicators (such as thyroid hormones and antibodies) and how these relationships manifest across different severities of OSA, sleep stages, and sex subgroups. After adjusting for multiple variables, the SASHB, NREM-SASHB, REM-SASHB, SBII, NREM-SBII, and REM-SBII were found to be independently associated with elevated FT3 levels in male patients, whereas no significant associations were observed in female subgroups.

Existing research highlights a bidirectional link between OSA and thyroid dysfunction. Hypothyroidism is prevalent in 25–35% of patients with OSA.³¹ However, the prevalence of OSA in patients with hypothyroidism is also high.^32,33 Our data align with this interplay: thyroid function indicators varied significantly across AHI groups, reinforcing the complex relationship between OSA severity and thyroid hormones. In previous studies conducted by our team, we reported that the progression of OSA in nonelderly individuals might promote an increase in thyroid hormone levels, particularly FT3, driven by oxidative stress and inflammation.¹⁷ This study further confirms this finding. Mechanistically, this interplay might can be explained by distinct responses to varying hypoxia severity. Mild hypoxia triggers compensatory thyroid hormone secretion via HPT axis activation. Conversely, severe hypoxia induces oxidative stress-mediated injury to thyroid follicular cells, impairing thyroid hormone synthesis.^12,13 Notably, most thyroid hormone levels in our cohort remained within normal ranges, potentially explaining discrepancies in prior studies.

The specific associations between SASHB, SBII, and thyroid function found in this study are novel. Existing research suggests that CIH in OSA can alter endocrine function by influencing thyroid hormone synthesis and release via the HPT axis, thereby affecting metabolic state.^14,34 Therefore, incorporating the depth and duration of CIH into thyroid function assessments is meaningful. Our analysis revealed that FT3 levels increased with higher SASHB and SBII quartiles, possibly due to hypoxia stimulating thyroid cells and affecting FT3 secretion. Regarding FT4, although there was a trend of differences in the SBII group comparisons, post hoc multiple comparisons revealed no significant differences, suggesting that the mechanisms by which FT4 is affected by OSA might be more complex, possibly involving other yet unidentified regulatory factors. Additionally, the trend of TSH levels initially decreasing and then increasing with SBII might indicate an early attempt to maintain thyroid stability via negative feedback, which becomes imbalanced as OSA progresses. Notably, a significant correlation was found between FT3 levels, SASHB and SBII, but not with AHI. This result strongly suggests that the SASHB and SBII capture key information that the AHI cannot encompass, thereby providing more accurate and sensitive insight into the intrinsic relationship between OSA and thyroid function.

The interplay between sleep stages and thyroid function was an interesting finding. Previous studies often focused on comparing the AHI during REM and NREM periods, classifying OSA populations into REM phenotypes and NREM phenotypes,^35,36 rather than studying the related sleep indicators for REM and NREM independently as a whole. Recently, more researchers have begun to recognize this, with one study finding a correlation between OSA during REM and NREM sleep and lipid levels.³⁷ These findings suggest that different sleep stages significantly impact endocrine and metabolic functions. In our study, the relationships between thyroid function indicators and OSA-related indicators differed during NREM and REM sleep. In NREM sleep, the FT3, FT4, and TSH levels significantly differed, with TSH initially decreasing but then increasing. This finding may be related to the physiological characteristics of NREM sleep. NREM sleep accounts for the majority of total sleep duration and is predominantly mediated by the parasympathetic nervous system,²⁷ during which metabolic and endocrine functions are relatively stable.²⁵ The deep stages of NREM sleep are particularly associated with the secretion of growth hormone,³⁸ significantly impacting thyroid function. Therefore, a greater SASHB and SBII during NREM sleep suggest that the patient experiences frequent or prolonged hypoxic and low oxygen saturation states during this phase, potentially leading to reduced secretion of thyroid hormones that might otherwise increase. In our study, only FT3 levels differed in relation to the REM-related SASHB and SBII groups, increasing with the severity of OSA. During REM sleep, increased sympathetic nervous system activity and a high metabolic state may suppress thyroid hormone secretion.^26,27 Thus, a higher SASHB and SBII during REM sleep indicate that frequent or prolonged hypoxia overrides typical REM-related thyroid hormone suppression, resulting in increased secretion of TH. In summary, considering the physiological changes associated with different sleep stages and their regulatory effects on thyroid function may provide a basis for personalized treatment for patients with OSA.

An unexpected finding was the absence of significant associations in the female subgroup, in contrast to male-specific correlations between SASHB, SBII and FT3. This disparity may stem from sex-specific hormonal influences on thyroid-hypoxia interactions. Estrogen can modulate thyroid hormone-binding globulin levels, subsequently affecting TH’s metabolism and activity.³⁹ Therefore, we speculate that during the pathophysiological changes associated with OSA, the hormonal dynamics could introduce variability that obscures direct OSA-TH associations in female. Additionally, the relatively small number of female patients in this study (57) may have limited the statistical power to detect associations between the AHI, SASHB, SBII, and thyroid function indicators, failing to adequately reflect the true physiological relationships.

This study is the first to investigate the relationships among the SASHB, the SBII, and thyroid function. Compared with the traditional AHI, these two indicators can more comprehensively reflect the characteristics of OSA-related respiratory events, capturing the degree of hypoxia and the duration and frequency of hypoxic events. By clearly establishing the independent correlations among FT3 and the SASHB and SBII, this study enriches our understanding of how OSA affects thyroid function and opens new avenues for research on the relationship between OSA and thyroid function. These findings suggest that future theoretical models should emphasize the impact of comprehensive quantitative indicators of sleep respiratory events on the endocrine system. Furthermore, unlike previous studies, we employed more refined sleep stage-related indicators to analyze their relationships with thyroid function indicators. Additionally, we conducted separate analyses for the male and female subgroups, revealing sex differences in the associations between OSA and thyroid function; specifically, multiple indicators in the male subgroup showed significant independent correlations with thyroid function, whereas no obvious associations were observed in the female subgroup, providing a reference for future targeted research on sex differences.

This study also has several limitations. First, as a retrospective study using historical medical records, information bias may be present. Second, although 452 participants with complete data were included, the sample size is still relatively small for an in-depth exploration of complex relationships, such as sex differences and the intricate relationships between different sleep stages and thyroid function and OSA indicators, which may limit the representativeness of the study results. Third, the single-center design introduces potential selection bias, affecting the generalizability of the findings. Notably, a large proportion of patients had severe OSA, likely reflecting clinical referral bias—patients with severe symptoms from a tertiary hospital sleep clinic were more likely to undergo PSG. Further multicenter, large-sample studies are needed to validate the conclusions of this study and improve its generalizability and reliability. Fourth, despite adjusting for potential confounders, unmeasured factors such as trace element deficiencies and medication history could still influence thyroid function, interfering with accurate assessment of the relationship between OSA and thyroid function. Fifth, this study is correlational and cannot determine the causal relationship between OSA and thyroid function abnormalities. Further prospective studies or animal experiments are needed for deeper exploration.

Conclusion

Overall, this study revealed that the SASHB and SBII are independently correlated with elevated FT3 levels in patients with OSA, with significant associations observed in male but not female subgroups and that this relationship varies across sleep stages. Future research needs to improve in aspects such as larger prospective sample sizes, multicenter collaborations, and optimized study designs to explore further the exact mechanisms of the impact of OSA on thyroid function, providing more valuable theoretical support for clinical diagnosis and treatment.

Abbreviations

AHI, apnea–hypopnea index; Anti-TG, thyroid globulin antibodies; Anti-TPO, thyroid peroxidase antibodies; BMI, body mass index; CHD, coronary heart disease; CT90, the percentage of sleep time with oxygen saturation below 90%; ESS, Epworth Sleepiness Scale; FT3, serum free triiodothyronine; FT4, serum free thyroxine; HPT, hypothalamic‒pituitary‒thyroid; Lowest SpO₂, lowest transcutaneous oxygen saturation at night; NC, neck circumference; NREM, non-rapid eye movement; NREM-SASHB, sleep apnea-specific hypoxic burden during non-rapid eye movement sleep; NREM-SBII, sleep breathing impairment index during non-rapid eye movement sleep; OSA, obstructive sleep apnea; PSG, polysomnography; REM, rapid eye movement; REM-SASHB, sleep apnea-specific hypoxic burden during rapid eye movement sleep; REM-SBII, sleep breathing impairment index during rapid eye movement sleep; RT3, reverse triiodothyronine; SASHB, sleep apnea-specific hypoxic burden; SBII, sleep breathing impairment index; T90, sleep time spent with oxygen saturation below 90%; TSH, serum thyroid stimulating hormone; TST, total sleep time; TT3, serum total triiodothyronine; TT4, secretes total thyroxine; WC, waist circumference.

Data Sharing Statement

The data supporting our findings are available on reasonable request from the corresponding author.

Ethics Approval

This retrospective study was approved by the Ethics Committee of the Second Affiliated Hospital of Xi’an Jiaotong University (approval no. 2022-1417). The procedures used in this study adhered to the tenets of the Declaration of Helsinki.

Acknowledgments

We wish to thank all who volunteered to participate in this study.

Author Contributions

YZ: conceptualization, methodology, software, writing original draft. YS: conceptualization, methodology, writing review and editing. SZ: software, formal analysis, writing original draft. ZC: investigation, resources, data curation, writing original draft. YX: investigation, resources. writing original draft CL: investigation, resources, writing review and editing. XN: investigation, resources, writing review and editing. LM: investigation, resources, writing review and editing. ZW: investigation, resources, writing original draft. YS: formal analysis, validation, writing original draft. ZX: formal analysis, validation, writing original draft. YY: formal analysis, validation, writing original draft. JY: formal analysis, validation, writing original draft. RL: formal analysis, writing original draft. YF: formal analysis, writing original draft. XR: conceptualization, methodology, supervision, funding acquisition, project administration, writing review and editing. WH: conceptualization, methodology, supervision, project administration, resources, writing review and editing. Xiaoyong Ren and Wei Hou are corresponding authors. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

This work was supported by the National Natural Science Foundation of China (grant no. 82371129) and the Free Exploration and Innovation Teacher Program of Xi’an Jiaotong University (no. xzy012023119). The funding bodies played no role in the study’s design, the collection, analysis, and interpretation of the data, or the writing of this paper.

Disclosure

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

1. Rundo JV. Obstructive sleep apnea basics. Cleve Clin J Med. 2019;86(9 Suppl 1):2–9. doi:10.3949/ccjm.86.s1.02

2. Jordan AS, Mcsharry DG, Malhotra A. Adult obstructive sleep apnoea. Lancet. 2014;383(9918):736–747. doi:10.1016/S0140-6736(13)60734-5

3. Gottlieb DJ, Punjabi NM. Diagnosis and management of obstructive sleep apnea: a review. JAMA. 2020;323(14):1389–1400. doi:10.1001/jama.2020.3514

4. Benjafield AV, Ayas NT, Eastwood PR, 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

5. Lyons MM, Bhatt NY, Pack AI, et al. Global burden of sleep-disordered breathing and its implications. Respirology. 2020;25(7):690–702. doi:10.1111/resp.13838

6. Framnes SN, Arble DM. The bidirectional relationship between obstructive sleep apnea and metabolic disease. Front Endocrinol. 2018;9:440. doi:10.3389/fendo.2018.00440

7. Mete T, Yalcin Y, Berker D, et al. Relationship between obstructive sleep apnea syndrome and thyroid diseases. Endocrine. 2013;44(3):723–728. doi:10.1007/s12020-013-9927-9

8. Mullur R, Liu YY, Brent GA. Thyroid hormone regulation of metabolism. Physiol Rev. 2014;94(2):355–382. doi:10.1152/physrev.00030.2013

9. Attal P, Chanson P. Endocrine aspects of obstructive sleep apnea. J Clin Endocrinol Metab. 2010;95(2):483–495. doi:10.1210/jc.2009-1912

10. Cheng SY, Leonard JL, Davis PJ. Molecular aspects of thyroid hormone actions. Endocr Rev. 2010;31(2):139–170. doi:10.1210/er.2009-0007

11. Hershman JM, Beck-Peccoz P. Discoveries around the hypothalamic-pituitary-thyroid axis. Thyroid. 2023;33(7):785–790. doi:10.1089/thy.2022.0258

12. Lavie L. Obstructive sleep apnoea syndrome–an oxidative stress disorder. Sleep Med Rev. 2003;7(1):35–51. doi:10.1053/smrv.2002.0261

13. Lavie L. Oxidative stress in obstructive sleep apnea and intermittent hypoxia–revisited–the bad ugly and good: implications to the heart and brain. Sleep Med Rev. 2015;20:27–45. doi:10.1016/j.smrv.2014.07.003

14. Lourbopoulos AI, Mourouzis IS, Trikas AG, et al. Effects of thyroid hormone on tissue hypoxia: relevance to sepsis therapy. J Clin Med. 2021;10(24):5855. doi:10.3390/jcm10245855

15. Ferreyra C, O’valle F, Osorio JM, et al. Effect of preconditioning with triiodothyronine on renal ischemia/reperfusion injury and poly(ADP-ribose) polymerase expression in rats. Transplant Proc. 2009;41(6):2073–2075. doi:10.1016/j.transproceed.2009.06.060

16. Xiong J, Hu H, Huang Y, et al. Lack of associations between thyroid dysfunction and obstructive sleep apnea-hypopnea syndrome: a meta-analysis. Medicine. 2023;102(49):e36531. doi:10.1097/MD.0000000000036531

17. Shi Y, Cao Z, Xie Y, et al. Association between obstructive sleep apnea and thyroid function: a 10-year retrospective study. Sleep Med. 2023;103:106–115. doi:10.1016/j.sleep.2023.01.027

18. Zhang M, Zhang W, Tan J, et al. Role of hypothyroidism in obstructive sleep apnea: a meta-analysis. Curr Med Res Opin. 2016;32(6):1059–1064. doi:10.1185/03007995.2016.1157461

19. Azarbarzin A, Sands SA, Stone KL, et al. The hypoxic burden of sleep apnoea predicts cardiovascular disease-related mortality: the osteoporotic fractures in men study and the sleep heart health study. Eur Heart J. 2019;40(14):1149–1157. doi:10.1093/eurheartj/ehy624

20. Cao W, Luo J, Huang R, et al. The association between sleep breathing impairment index and cardiovascular risk in male patients with obstructive sleep apnea. Nat Sci Sleep. 2022;14:53–60. doi:10.2147/NSS.S343661

21. Malhotra A, Ayappa I, Ayas N, et al. Metrics of sleep apnea severity: beyond the apnea-hypopnea index. Sleep. 2021;44(7). doi:10.1093/sleep/zsab030

22. Li C, Peng Y, Zhu X, et al. Independent relationship between sleep apnea-specific hypoxic burden and glucolipid metabolism disorder: a cross-sectional study. Respir Res. 2024;25(1):214. doi:10.1186/s12931-024-02846-7

23. Dai L, Cao W, Luo J, et al. The effectiveness of sleep breathing impairment index in assessing obstructive sleep apnea severity. J Clin Sleep Med. 2023;19(2):267–274. doi:10.5664/jcsm.10302

24. Azarbarzin A, Sands SA, Taranto-Montemurro L, et al. The sleep apnea-specific hypoxic burden predicts incident heart failure. Chest. 2020;158(2):739–750. doi:10.1016/j.chest.2020.03.053

25. Aserinsky E, Kleitman N. Regularly occurring periods of eye motility, and concomitant phenomena, during sleep. Science. 1953;118(3062):273–274. doi:10.1126/science.118.3062.273

26. España RA, Scammell TE. Sleep neurobiology from a clinical perspective. Sleep. 2011;34(7):845–858. doi:10.5665/SLEEP.1112

27. De Zambotti M, Baker FC. Sleep and circadian regulation of the autonomic nervous system. Autonomic Nervous System Sleep. 2021;63–69.

28. Alzoubaidi M, Mokhlesi B. Obstructive sleep apnea during rapid eye movement sleep: clinical relevance and therapeutic implications. Curr Opin Pulm Med. 2016;22(6):545–554. doi:10.1097/MCP.0000000000000319

29. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14(6):540–545. doi:10.1093/sleep/14.6.540

30. Sateia MJ. International classification of sleep disorders-third edition: highlights and modifications. Chest. 2014;146(5):1387–1394. doi:10.1378/chest.14-0970

31. Kapur VK, Koepsell TD, Demaine J, et al. Association of hypothyroidism and obstructive sleep apnea. Am J Respir Crit Care Med. 1998;158(5 Pt 1):1379–1383. doi:10.1164/ajrccm.158.5.9712069

32. Lin CC, Tsan KW, Chen PJ. The relationship between sleep apnea syndrome and hypothyroidism. Chest. 1992;102(6):1663–1667. doi:10.1378/chest.102.6.1663

33. Grunstein RR, Sullivan CE. Sleep apnea and hypothyroidism: mechanisms and management. Am J Med. 1988;85(6):775–779. doi:10.1016/S0002-9343(88)80020-2

34. Fisher DA. Physiological variations in thyroid hormones: physiological and pathophysiological considerations. Clin Chem. 1996;42(1):135–139. doi:10.1093/clinchem/42.1.135

35. Karuga FF, Kaczmarski P, Białasiewicz P, et al. REM-OSA as a tool to understand both the architecture of sleep and pathogenesis of sleep apnea-literature review. J Clin Med. 2023;12(18):5907. doi:10.3390/jcm12185907

36. Joosten SA, Landry SA, Wong AM, et al. Assessing the physiologic endotypes responsible for REM- and NREM-based OSA. Chest. 2021;159(5):1998–2007. doi:10.1016/j.chest.2020.10.080

37. Xu H, Xia Y, Li X, et al. Association between obstructive sleep apnea and lipid metabolism during REM and NREM sleep. J Clin Sleep Med. 2020;16(4):475–482. doi:10.5664/jcsm.8242

38. Honda Y, Takahashi K, Takahashi S, et al. Growth hormone secretion during nocturnal sleep in normal subjects. J Clin Endocrinol Metab. 1969;29(1):20–29. doi:10.1210/jcem-29-1-20

39. Marqusee E, Braverman LE, Lawrence JE, et al. The effect of droloxifene and estrogen on thyroid function in postmenopausal women. J Clin Endocrinol Metab. 2000;85(11):4407–4410. doi:10.1210/jcem.85.11.6975