The Influence of Comorbidities on Chemokine and Cytokine Profile in Obstructive Sleep Apnea Patients: Preliminary Results

Introduction: Obstructive sleep apnea (OSA) is frequently associated with a chronic inflammatory state and cardiovascular/metabolic complications. The aim of this study was to evaluate the influence of certain comorbidities on a panel of 45 chemokines and cytokines in OSA patients with special regard to their possible association with cardiovascular diseases. Material and Methods: This cross-sectional study was performed on 61 newly diagnosed OSA patients. For the measurement of the plasma concentration of chemokines and cytokines, the magnetic bead-based multiplex assay for the Luminex® platform was used. Results: In the patients with concomitant COPD, there were increased levels of pro-inflammatory cytokines (CCL11, CD-40 ligand) and decreased anti-inflammatory cytokine (IL-10), while in diabetes, there were increased levels of pro-inflammatory cytokines (IL-6, TRIAL). Obesity was associated with increased levels of both pro-inflammatory (IL-13) and anti-inflammatory (IL-1RA) cytokines. Hypertension was associated with increased levels of both pro-inflammatory (CCL3) and anti-inflammatory (IL-10) cytokines. Increased daytime pCO2, low mean nocturnal SaO2, and the oxygen desaturation index were associated with increased levels of pro-inflammatory cytokines (CXCL1, PDGF-AB, TNF-α, and IL-15). Conclusions: In OSA patients with concomitant diabetes and COPD, elevated levels of certain pro-inflammatory and decreased levels of certain anti-inflammatory cytokines may favor the persistence of a chronic inflammatory state with further consequences. Nocturnal hypoxemia, frequent episodes of desaturation, and increased daytime pCO2 are factors contributing to the chronic inflammatory state in OSA patients.


Introduction
Obstructive sleep apnea (OSA) is the most important sleep-breathing disorder of clinical significance. The occurrence of OSA ranges from 10% to 17% in men and from 3% to 9% in women-more frequently in older (≥50 years) persons [1]. Recurrent episodes of sleep apneas and hypopneas cause episodes of arterial oxygen desaturation and, in consequence, lead to oxidative stress, endothelial dysfunction, neurohormonal dysregulation, sleep fragmentation, and, finally, changes in the central nervous and cardiovascular systems [2][3][4][5][6][7].
Common comorbidities associated with OSA include obesity, hypertension, diabetes mellitus, and chronic obstructive pulmonary disease (COPD) [8,9]. OSA is also a risk factor of atrial fibrillation for the recurrence of atrial fibrillation after cardioversion and/or ablation [10] and for other arrhythmias [11]. OSA favors the occurrence of left heart failure [12,13] and, in some patients, pulmonary hypertension [14]. In severe cases, the risk

Patients and Controls
In this cross-sectional study, 61 patients (F/M = 10/51) with OSA syndrome were investigated. Inclusion criteria were as follows: Age > 40 years, diagnosis based on inhospital polysomnography (PSG), and no previous OSA treatment. Exclusion criteria encompassed a lack of agreement for participation in the study and the absence of any unstable or acute disease.
There were 42 obese (body mass index, BMI > 30 kg/m 2 ) patients (69%) and 19 nonobese patients (31%), including 16 overweight patients and 3 normal-weight patients. There were 7 patients (11%) with overlap syndrome, 16 patients (26%) with diabetes mellitus, and 43 patients (70%) with arterial hypertension. The comparison of clinical data in the groups of patients with and without comorbidities is shown in Table 1. Hypertensive patients were older than normotensive patients, diabetic patients were more obese than non-diabetic patients, overlap patients had lower mean nocturnal SaO 2 and lower daytime pO 2 than the patients without COPD, and obese patients had higher AHI and lower daytime pO 2 than non-obese patients. All participants in the study provided written informed consent. The study was approved by the local Ethics Committee (No 1082/2021), and all the procedures were in accordance with the ethical standards of the Helsinki Declaration, as revised in 2013.

Polysomnography
All the patients underwent in-hospital polysomnography using the Alice 6 LDe Polysomnographic Sleep System (Philips Respironics, Monroeville, PA, USA). During 8 h of nocturnal sleep, the following parameters were measured: Airflow with the use of an oronasal thermal sensor and a nasal pressure sensor, chest and abdomen movements, oxygen saturation using a finger clip sensor for respiration, and electroencephalography, electromyography, and electrooculography for sleep stages. The following parameters were analyzed: AHI, ODI, mean arterial oxygen saturation (SaO 2 ) during sleep, and minimal SaO 2 at the end of sleep apnea/hypopnea episodes. Apneas were defined as the complete cessation of airflow for >10 s with concomitant respiratory movements of the chest and diaphragm, and hypopneas were defined as 30-50% reduction in oronasal airflow for >10 s associated with desaturation >3% or with arousal. Manual scoring was carried out after automatic scoring, according to the American Association of Sleep Medicine criteria [36]. OSA syndrome diagnosis was based on AHI >5/h and the presence of symptoms such as excessive daytime somnolence and daytime fatigue with concomitant choking and recurrent awakenings during sleep.

Chemokine and Cytokine Serum Levels
Serum samples were collected using BD Vacutainer SSTTM II Advance tubes (Becton Dickinson, Franklin Lakes, NJ, USA) from all participants and stored at −20 • C. Samples were then thawed and screened for the simultaneous detection of 45 chemokines and cytokines with the use of a customized Human XL Cyt Disc Premixed Mag Luminex Perf Assay Kit (R&D Systems Inc., Minneapolis, MN, USA). Analyses were performed according to the manual provided by the manufacturer. Serum samples were not diluted for the experiment. For analysis purposes, the Luminex 200 instrument (Luminex Corp., Austin, TX, USA) was used.
The A list of the cytokines and chemokines studied with official names and gene locus is presented in Table 2. A list of the cytokines/chemokines with their function related to inflammation and cardiovascular diseases  is presented in Table 3. There were no valid experimental results in the case of the following 7 cytokines/chemokines: B7-H1/PD-L1, IFN-ß, IL-3, IL-12 p70, IL-17A, IL-17E/IL-25, and IL-33, which is why these cytokines/chemokines were not analyzed in further statistical analysis.   Table 3. The list of cytokines/chemokines with their function related to inflammation and cardiovascular diseases.

Statistical Analysis
Results of the serum cytokine and chemokine level assessment were related to clinical data. All statistical analyses were performed using STATISTICA 13 software (StatSoft. Inc., Tulsa, OK, USA). The U Mann-Whitney test for two independent samples was used, and for correlation analyses, the Spearman's Rho correlation test was used. In all calculations, the statistical significance was considered at p < 0.05. The effect size was measured as either the standardized mean difference (β) or Spearman's Rho correlation coefficient (rs).

Results
The comparison of the concentrations of the chemokines and cytokines in the groups of patients with and without comorbidities is shown in Table 4. In obese patients, there were increased concentrations of IL-1ra and IL-13 compared with non-obese patients. In overlap patients, there were increased concentrations of CCL11 and CD40 ligands but decreased concentrations of IL-10 compared with patients without concomitant COPD.  In hypertensive patients, there were increased concentrations of CCL3 and IL-10 compared with normotensive patients. In diabetic patients, there were increased concentrations of IL-6 and TRAIL compared with non-diabetic patients.
The comparison of OSA patients without any other diseases with OSA patients with only hypertension also demonstrated increased concentrations of CCL3 and, additionally, increased concentrations of CX3CL1/Fractalkine and IL-7 (Table 5).  Correlations between chemokine/cytokine levels and BMI, daytime pO 2 and pCO 2 , and parameters indicating the severity of OSA (AHI, ODI, mean SaO 2 , and minimal SaO 2 ) are shown in Table 6. There was a positive correlation between BMI and IL-1ra, and IFN-γ was negatively correlated with BMI. There was a positive correlation between daytime pO 2 and TRAI and an inverse correlation between pCO 2 and TRAIL. Daytime pO2 also positively correlated with the concentration of CCL11 and negatively correlated with the concentration of IL-1ra; daytime pCO 2 was positively correlated with the concentration of CXCL1 and PDGF-AB. There was a positive correlation between ODI and TNF-α, as well as between mean SaO 2 during sleep and IL-15, with no influences of AHI or minimal SaO 2 at the end of sleep apneas/hypopneas.

Discussion
There were differences in the concentrations of some chemokines/cytokines in patients with and without obesity, COPD, hypertension, and diabetes mellitus, as well as differences associated with daytime gas exchange and sleep hypoxemia.
Obesity had some influence on the cytokine/chemokine profile as shown by higher concentrations of IL-1ra and IL-13 in obese compared to non-obese OSA patients. In addition, there was a positive correlation between IL-1RA and BMI. IL-1RA has antiinflammatory properties [82]. In the previous studies, obese, otherwise healthy persons also had elevated IL-1RA levels [72]. In patients with rheumatoid disease, IL-1ra concentrations positively correlated with BMI [67]. In the OSA patients, IL-1ra levels were also increased, and weight loss resulted in a decrease in its expression [82]. Increased levels of IL-1RA in obese OSA patients may be regarded as a protective factor. This may also confirm the negative correlation between IFN-γ and BMI. IFN-γ belongs to pro-inflammatory cytokines [83]. In adult OSA patients, elevated levels of IFNγ were found in the group with concomitant coronary heart disease [84]. In children with OSA, IFN-γ negatively correlated with cardiac function [85].
Overlap syndrome was associated with increased levels of the CCL11 and CD40 ligands, as well as decreased IL-10 levels. In COPD patients, there is a broad dysregulation of chemokines, including-as seen in our study in overlap patients-increased levels of CCL11 [86]. CCL11 has pro-inflammatory properties [87].
Increased levels of the CD40 ligand were observed in COPD patients, negatively correlating with ventilatory impairment [88]. This cytokine belongs to pro-inflammatory cytokines [89].
Decreased concentrations of Il-10 in stable COPD patients were found [90]. This is in line with the pathogenesis of COPD, as IL-10 is an anti-inflammatory cytokine [91]. In obese COPD patients, the levels of IL-10 were not decreased, indicating more severe inflammation than in non-obese COPD patients [92]. However, in some studies, COPD patients had elevated Il-10 levels [93].
In our OSA patients with concomitant arterial hypertension, there were increased concentrations of CCL3 and IL-10. In addition, the comparison of OSA patients without any other disease with OSA patients with only hypertension also demonstrated increased levels of CXCL2/GRO-ß and IL-7. In other studies, in children with primary hypertension, the serum levels of CCL3 were not different than in normotensive children [94]. Circulating levels of CCL3 (as well as CXCL10 and CD40 ligands) were associated with heart failure [50]. CCL3 has pro-inflammatory properties [95]. The role of inflammation in hypertension is still incompletely explained. There are numerous associations between changes in blood pressure and inflammatory mediators, including-as seen in our study-IL-10 and IL-7, but also interferon-γ, GM-CSF, IL-4, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, IL-17A, IL-21, IL-23, MIP-1α, and MIP-1β [96]. CXCL-2 belongs to the cytokines involved in the pathogenesis of some cardiovascular diseases as acute myocardial infarction, atherosclerosis, obesity, diabetes, and ischemic stroke [40].
Decreased levels of IL-10 were observed in OSA patients [97]. Differences in IL-10 concentrations in the OSA patients in relation to the presence of hypertension were found, with lower IL-10 concentrations in hypertensive than in normotensive patients [98]. In hypertensive patients, IL-10 was decreased [99]. IL-10 belongs to anti-inflammatory cytokines [100].
The influence of diabetes on the cytokine profile in our OSA patients was shown by increased levels of IL-6 and TRAIL. The data on IL-6 concentrations in OSA patients are contradictory. IL-6 was found to be increased in OSA patients compared to non-OSA patients, either obese or non-obese [101]. In another study, no significant differences in IL-6 were found in OSA and non-OSA patients [102]. In patients with OSA and diabetes, increased IL-6 levels were observed [71]. There is an influence of obesity on IL-6 levels, as its levels are positively correlated with BMI [103]. An increased IL-6 level was an independent predictor of type 2 diabetes and played an important role in inflammation, insulin resistance, and beta-cell dysfunction [104]. IL-6 belongs to pro-inflammatory cytokines [97].
TRAIL induces an inflammatory response, which stimulates the expression of chemokines and cytokines, including IL-6 [105]. Diabetic patients had decreased TRAIL levels compared to healthy controls [106,107]. There is a negative correlation between TRAIL levels and cardiovascular risk [80]. Previous studies indicated the importance of TRAIL in the development and progression of diabetes [108]. It is implicated in the regulation of T cell activation and suppresses the inflammatory process in many autoimmune diseases [109]. This modulation of the immune system also protects against diabetes [108]. TRIAL has pro-inflammatory properties [110].
In our study, in OSA patients, there was a positive correlation between TRAIL and daytime pO 2 , as well as an inverse correlation between TRAIL and daytime pCO 2 . TRAIL is a factor involved in the development of pulmonary hypertension [111,112]-a condition that may be a consequence of alveolar hypoxia and chronic hypoxemia. In OSA patients, repetitive sleep apneas and hypopneas cause alveolar hypoxia, and in cases with chronic daytime hypoxemia, pulmonary hypertension develops [113].
There was a positive correlation between pO 2 and the levels of both IL-1ra and CCL11, indicating an increase in both anti-inflammatory and pro-inflammatory actions along with improved daytime gas exchange in OSA patients.
On the other hand, however, increasing daytime pCO 2 was positively correlated with the level of CXCL1 and PDGF-AB. As both CXCL1 [114] and PDGF-AB [115] have pro-inflammatory properties, this observation indicates that the tendency to hypoventilation, as shown by increasing pCO 2 , is associated with increased inflammatory status in OSA patients.
There was a positive correlation between IL-15 and mean nocturnal SaO 2 . IL-15 belongs to pro-inflammatory cytokines [116], which indicates the association between sleep hypoxemia as a factor contributing to a chronic inflammatory state. This also confirms the positive correlation between ODI and TNF-α. TNF-α belongs to pro-inflammatory cytokines [117], which play an important role in OSA. Its serum levels increase with OSA severity and correlate with the frequency of apnea and hypopnea [118].
Our study has certain limitations. First, we divided the group of OSA patients into subgroups and compared the subgroups of OSA patients with and without comorbidities, i.e., obesity, hypertension, diabetes mellitus, and COPD, but did not compare these comorbidities with control groups. Second, the compared groups were relatively small. Moreover, the subgroups were not "pure", e.g., all diabetic OSA patients also had hypertension, and obesity was diagnosed in 69% of our patients, thus some obese patients had to be included in other subgroups.
The main strength of the study is that in our OSA patients with the most common comorbidities, a significantly higher number of cytokines/chemokines were concomitantly measured. To the best of our knowledge, such an extensive study on cytokines and chemokines in OSA patients has not been performed.
In summary, the chemokine/cytokine profile in OSA patients with concomitant diseases indicates the inflammatory status in overlap syndrome, as shown by increased levels of pro-inflammatory proteins (CCL11, CD-40 ligand) and decreased anti-inflammatory protein (IL-10), and in diabetes, as shown by increased levels of pro-inflammatory cytokines (IL-6, TRIAL). There is an increase in the levels of both pro-inflammatory and anti-inflammatory cytokines in OSA patients with obesity (IF-γ and IL-1RA, respectively) or hypertension (CCL3 and IL-10, respectively). Increasing daytime pCO 2, low mean nocturnal SaO 2 , and ODI are associated with increased levels of pro-inflammatory cytokines (CXCL1, PDGF-AB, IL-15, and TNF-α, respectively).

Conclusions
In OSA patients with concomitant diabetes and COPD, elevated levels of certain pro-inflammatory and decreased levels of certain anti-inflammatory cytokines may favor the persistence of a chronic inflammatory state with further consequences. Nocturnal hypoxemia, frequent episodes of desaturation, and increased daytime pCO 2 are the factors contributing to the chronic inflammatory state in OSA patients.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy.

Conflicts of Interest:
The authors declare no conflict of interest.