Association of Tumor Necrosis Factor-Alpha, Interleukin-1β, Interleukin-8, and Interferon-γ with Obstructive Sleep Apnea in Both Children and Adults: A Meta-Analysis of 102 Articles

Background: Cytokines may have a significant impact on sleep regulation. In this meta-analysis, we present the serum/plasma levels of tumor necrosis factor-alpha (TNF-α), interleukin (IL)-8, IL-1β, and interferon-gamma (IFN-γ) in both children and adults with obstructive sleep apnea (OSA) in comparison to controls. Methods: Four electronic databases were systematically searched (PubMed, Web of Science, Scopus, and Cochrane Library) through 19 October 2023, without any restrictions on language, date, age, and sex. We used Review Manager version 5.3 to perform meta-analysis and presented the data as standardized mean difference (SMD) and 95% confidence interval (CI) values to evaluate the relationships between the levels of cytokines and OSA. Results: A total of 102 articles (150 independent studies) were included in the meta-analysis. The pooled SMDs in adults were 1.42 (95%CI: 1.11, 1.73; p < 0.00001), 0.85 (95%CI: 0.40, 1.31; p = 0.0002), 0.69 (95%CI: 0.22, 1.16; p = 0.004), and 0.39 (95%CI: −0.37, 1.16; p = 0.31) for TNF-α, IL-8, IL-1β, and IFN-γ, respectively. The pooled SMDs in children were 0.84 (95%CI: 0.35, 1.33; p = 0.0008), 0.60 (95%CI: 0.46, 0.74; p < 0.00001), 0.25 (95%CI: −0.44, 0.93; p = 0.49), and 3.70 (95%CI: 0.75, 6.65; p = 0.01) for TNF-α, IL-8, IL-1β, and IFN-γ, respectively. Conclusions: The levels of proinflammatory cytokines of TNF-α, IL-8, and IL-1β in adults, and TNF-α, IL-8, and IFN-γ in children with OSA, are significantly higher than those in controls.


Introduction
Obstructive sleep apnea (OSA) is a common disorder characterized by episodes of partial or complete upper airway obstruction during sleep, often resulting in disrupted sleep and other health issues [1,2].In the general population, nearly 1 billion adults aged 30-69 years worldwide have mild-to-severe OSA, and 425 million people worldwide have moderate-to-severe OSA [3].Polysomnography is the standard diagnostic test for OSA in both adults and children [4].The apnea-hypopnea index (AHI), where mild OSA has an AHI from 5 to <15, moderate has an AHI from 15 to 30, and severe has an AHI of >30, with this value representing the average number of apneas and hypopneas that occur per hour of sleep, is quantified during this test [5].
The occurrence of OSA in adults varies.It is estimated that about 15 to 30 percent of males and 10 to 15 percent of females have OSA when it is broadly defined as an AHI of 5 or more events/h [6][7][8].However, when stricter definitions are applied (such as AHI ≥ 5 events/h with symptoms or AHI ≥ 15 events/h), the estimated prevalence drops to around 15% and 5% in males and females, respectively [6,7,9].In children, the prevalence of OSA is estimated to range from 0.7% to 13% [10].This broad range can be attributed to

Study Selection
The PICOS framework was introduced as the inclusion criteria (Population (P): both children and adults diagnosed with OSA.Intervention (I): measurement of TNF-α, IL-8, IL-1β, and IFN-γ levels.Comparison (C): comparison of these levels with a control group of both children and adults without OSA.Outcome (O): the association of TNF-α, IL-8, IL-1β, and IFN-γ levels with OSA.Study design (S): observational studies).OSA patients were diagnosed according to the clinical practice guidelines from the American College of Physicians used for adults [36] and children [37].The levels of cytokines were determined using pg/mL or calculated in relation to the accordant unit.
The inclusion criteria involved: (1) Studies that included adult participants with an AHI ≥ 5 events/h as adults with OSA and AHI < 5 events/h as adults without OSA and child participants with an AHI ≥ 1 events/h as children with OSA and AHI < 1 events/h as children without OSA.(2) Studies that used morning levels of cytokines.(3) OSA was diagnosed via polysomnography.
The exclusion criteria included the following: (1) Participants with a history or diagnosis of any systemic diseases overlapping with OSA (e.g., other respiratory, cardiovascular, and endocrine diseases); patients who received nutritional support or underwent therapy (e.g., medication, operation, continuous positive airway pressure); case reports or those articles lacking statistical data; and articles without a control group.(2) Articles selected an adult control group with an AHI > 5 events/h or a child control group with an AHI > 1 event/h.(3) Reviews, meta-analyses, letters to the editors, and book chapters.(4) Articles without any data; articles which measured the level of cytokines in saliva, urine, exhaled breath condensate, and cells; and articles which measured the overnight or evening levels of cytokines.( 5) Gene expression and a high sensitivity level of cytokines.

Data Extraction
Two authors (M.S. and E.S.) independently screened the literature and extracted data to ensure that the screening core criteria and data gathering were consistent.If the opinions were different, further discussion was carried out or an additional person (A.G.) was invited to participate in discussions until a consensus was reached.We extracted the first author, year of publication, nationality, sample size, gender, mean BMI, mean age, mean AHI, blood sample, and levels of cytokines.The quality of the included studies was evaluated according to the Newcastle-Ottawa scale (NOS) [38].Two authors independently performed the process of screening title/abstract and screening full texts, and a consensus was reached on all decisions.

Statistical Analysis and Data Synthesis
We used Review Manager version 5.3 for meta-analysis and presented the data as standardized mean difference (SMD) and 95% confidence interval (CI) in order to evaluate the relationship between the levels of cytokines and OSA.We utilized GetData Graph Digitizer software version 2.26 to report mean ± SD on the graphs.In some studies, median (interquartile) or median (range) was reported, which we converted into mean ± standard deviation (SD).We converted standard error (SE) into SD by SE = SD √ N , N = Number of cases.We extracted mean ± SD from log (mean ± SD) in "https://brilliant.org/wiki/log-normaldistribution/",accessed on 30 January 2024.
The heterogeneity of the studies was assessed using the I2 statistic, with the level of significance set at p < 0.05.Given the probable heterogeneity of the studies, a randomeffects model was employed in the meta-analysis when P heterogeneity was less than 0.10 (I 2 more than 50%) [39].Otherwise, a fixed-effect model was used [40].Publication bias was evaluated using a funnel plot and Begg's and Egger's tests, with the level of significance set at p < 0.10.Comprehensive meta-analysis version 2.0 (CMA 2.0) software was used to perform bias analyses, meta-regression, and sensitivity analyses.
We conducted a trial sequential analysis (TSA) using TSA software (version 0.9.5.10 beta) [41].The required information size (RIS) was calculated for blood cytokine levels, with alpha risk set at 5% and beta risk at 20%.The mean difference was based on empirical assumptions.If the Z-curve interrupted the RIS, sufficient cases were entered into the studies and the conclusion could be reliable.

Search Strategy
Records were identified through database searching (1876 records) and the use of other sources (10 records).After removing duplicates, there were 1059 unique records, all of which were screened.Full-text articles were assessed for eligibility (181 articles).However, due to various reasons, only 102 articles were included in the systematic review.All of these 102 articles [8,16,21,[31][32][33]] also appeared to be included in a meta-analysis (Figure 1).

Search Strategy
Records were identified through database searching (1876 records) and the use of other sources (10 records).After removing duplicates, there were 1059 unique records, all of which were screened.Full-text articles were assessed for eligibility (181 articles).However, due to various reasons, only 102 articles were included in the systematic review.All of these 102 articles [8,16,21,[31][32][33]] also appeared to be included in a meta-analysis (Figure 1).

Subgroup Analysis
Table 3 shows the subgroup analysis of several variables affecting TNF-α, IL-8, and IL-1β levels in adults and TNF-α levels in children with OSA compared to controls.The results recommended that ethnicity was an effective factor for all factors, and also that blood samples and mean BMI played roles for IL-8 and IL-1β levels in adults, for sample size, mean age, and mean AHI for IL-8 and IL-1β in adults, and for TNF-α levels in children.

Subgroup Analysis
Table 3 shows the subgroup analysis of several variables affecting TNF-α, IL-8, and IL-1β levels in adults and TNF-α levels in children with OSA compared to controls.The results recommended that ethnicity was an effective factor for all factors, and also that blood samples and mean BMI played roles for IL-8 and IL-1β levels in adults, for sample size, mean age, and mean AHI for IL-8 and IL-1β in adults, and for TNF-α levels in children.

Subgroup Analysis
Table 3 shows the subgroup analysis of several variables affecting TNF-α, IL-8, and IL-1β levels in adults and TNF-α levels in children with OSA compared to controls.The results recommended that ethnicity was an effective factor for all factors, and also that blood samples and mean BMI played roles for IL-8 and IL-1β levels in adults, for sample size, mean age, and mean AHI for IL-8 and IL-1β in adults, and for TNF-α levels in children.

Subgroup Analysis
Table 3 shows the subgroup analysis of several variables affecting TNF-α, IL-8, and IL-1β levels in adults and TNF-α levels in children with OSA compared to controls.The results recommended that ethnicity was an effective factor for all factors, and also that blood samples and mean BMI played roles for IL-8 and IL-1β levels in adults, for sample size, mean age, and mean AHI for IL-8 and IL-1β in adults, and for TNF-α levels in children.

Subgroup Analysis
Table 4 shows a random meta-regression analysis for cytokines, with seven variables used.For IL-8 in adults, the coefficient is 0.0737 for mean AHI in these cases, indicating that for each unit increase in AHI, the expected change in IL-8 is an increase of 0.0737 units.In these cases, there is a 0.2613 mean BMI in controls, indicating that for each unit increase in BMI in controls, the expected change in IL-8 is a decrease of 0.2613 units.There is a 0.1160 mean age in controls, indicating that for each unit increase in age in controls, the expected change in IL-8 is an increase of 0.1160 units.The p-value is less than 0.05, indicating that these effects are statistically significant.
For TNF-α in children, the coefficient is 0.0524 for mean AHI in cases, indicating that for each unit increase in AHI in cases, the expected change in TNF-α is an increase of 0.0524 units when holding all other variables constant.The p-value is 0.0314, which is less than 0.05, indicating that this effect is statistically significant.
For IL-8 in children, all variables in this category are statistically significant, with p-values less than 0.05.This means that all these variables, namely publication year, sample size, mean BMI in cases, mean age in cases, mean AHI in cases, mean BMI in controls, and mean age in controls, are confounding factors for the IL-8 biomarker in children's cases.

Publication Bias
Table 5 presents the results of the publication bias analysis for various biomarkers in both adult and child populations.The analysis was conducted using two tests: Egger's test and Begg's test.A p-value of less than 0.10 is considered statistically significant and indicates potential publication bias.In this case, the biomarker of TNF-α (adult) shows significant publication bias in both tests, with p-values less than 0.0001.The biomarker of IL-8 (children) also shows potential publication bias in Egger's test, with a p-value of 0.0507, but not in Begg's test.All other biomarkers do not show statistically significant publication bias in either of the tests.Supplementary File S2 shows the funnel plots.

Sensitivity Analysis
The pooled results were stable for four cytokines in both adults and children, because "one-study-removed" and "cumulative" analyses did not change the results.

TSA
The Z-curve did not cross the RIS for IFN-γ in adults and IL-1β in children, indicating that the evidence is not yet sufficient to confirm an effect and that more trials are needed.However, for the other biomarkers, the Z-curve crossed the RIS, suggesting that enough evidence was accumulated to make conclusions.Please refer to Supplementary File S3 for the specific TSA plots.

Discussion
The meta-analysis presents a comparison between blood levels of TNF-α, IL-8, IL-1β, and IFN-γ in OSA in adults and children, comparing patients with OSA and healthy controls.The results show higher levels of TNF-α, IL-8, and IL-1β in adults with OSA, and higher levels of TNF-α, IL-8, and IFN-γ in children with OSA compared to controls.Subgroup analysis revealed that ethnicity, blood sample, mean BMI, sample size, mean age, and mean AHI are significant factors influencing these levels.Meta-regression analysis further confirmed these findings, with all variables showing statistical significance for IL-8 in children.Publication bias analysis indicated a potential bias for TNF-α in adults and IL-8 in children.The TSA suggested that more trials are needed for IFN-γ in adults and IL-1β in children, while sufficient evidence has been gathered for the other biomarkers.
The development of OSA is a complex process involving multiple factors, one of which includes the selective activation of inflammatory response pathways [134].Studies have shown that intermittent hypoxia, a condition often seen in OSA, can trigger inflammatory pathways, leading to cardiovascular or metabolic diseases [135].Factors such as oxidative stress, cardiovascular inflammation, endothelial dysfunction, and metabolic abnormalities in OSA could hasten the process of atherogenesis [136].The regulation of systemic and airway inflammation in OSA could vary depending on the type of cytokine, and could be influenced by OSA itself and concurrent obesity [137].
In cases of OSA, there is an observed increase in the levels of proinflammatory cytokines and a decrease in anti-inflammatory factors, ultimately leading to endothelial dysfunction [138].The current meta-analysis identifies all the mentioned cytokines, namely TNF-α, IL-8, IL-1β, and IFN-γ, as proinflammatory cytokines [139].The results indicate elevated levels of TNF-α, IL-8, and IL-1β in adults with OSA, and increased levels of TNF-α, IL-8, and IFN-γ in children with OSA when compared to control groups.These cytokines are pivotal in regulating immune responses and inflammation.
A review highlighted that patients with OSA are more prone to coronary atherosclerosis and to exhibit signs of cardiac remodeling and dysfunction [140].In 2020, another study affirmed that OSA could be viewed as a systemic inflammatory disease, with an imbalance of non-neuronal cholinergic and pro/anti-inflammatory cytokines playing a role in the onset and progression of comorbidities in OSA patients [141].A 2020 review discovered that the recurring episodes of airway collapse and obstruction in OSA patients lead to apnea and arousal during sleep, which results in intermittent hypoxia and excessive daytime sleepiness, thereby contributing to inflammation [142].These studies underscore the connection between OSA and inflammation, suggesting that the effective management of OSA could help to mitigate inflammation and enhance patient outcomes.However, the precise mechanisms are intricate and continue to be the subject of ongoing research.
A meta-analysis by Imani in 2020 [20] found that factors such as ethnicity, BMI, AHI, age, and sample size significantly influenced the blood levels of TNF-α in OSA patients compared to control groups.Two other meta-analyses highlighted that the severity of OSA also impacts the blood levels of TNF-α [24,26].Another meta-analysis by Li et al. [30] indicated that age and ethnicity play a role in the relationship between OSA and IL-8 levels.The current meta-analysis reaffirms that factors like ethnicity, age, BMI, AHI, and sample size have varying effects on different cytokines.These variations could be attributed to the unique mechanisms of each cytokine, which can be further investigated in the future.The present meta-analysis reported that as the AHI score increased, the IL-8 levels in adults and children and TNF-α level in children significantly increased.As the age increased, the IL-8 levels in adults and children significantly increased.In addition, as the sample size and BMI increased, the IL-8 levels in children significantly increased.The results suggest that the levels of cytokines, which are small proteins important in cell signaling, vary among different ethnicities.It is important to note that many factors can contribute to these differences, including genetic factors, environmental influences, lifestyle choices, and more.It is also crucial to remember that while such research can identify trends across populations, individual health outcomes can vary widely within any ethnic group.Therefore, these findings should be interpreted with caution and used as a starting point for further research, rather than as definitive conclusions about individual health.

Strengths
(1) This study includes a systematic review of 102 articles, providing a broad overview of the research in this field.(2) This study conducts subgroup analysis for TNF-α, IL-8, and IL-1β levels in adults and TNF-α levels in children with OSA compared to controls.This allows for a more nuanced understanding of the effects of these variables.(3) This study performs meta-regression analysis for cytokines with seven variables, providing insights into the relationships between these variables and the cytokines.(4) This study conducts a publication bias analysis, ensuring the robustness of the results.(5) Sensitivity analyses reports the stability of pooled results.

Limitations
(1) Heterogeneity was high for most of the analyses.This suggests that there was considerable variability in the results of the included studies.(2) The publication bias analysis showed significant bias for the biomarker TNF-α in adults and potential bias for the biomarker IL-8 in children.This could affect the reliability of the results.(3) The Z-curve did not cross the RIS for IFN-γ in adults and IL-1β in children, indicating that the evidence is not yet sufficient to confirm an effect and that more trials are needed.(4) Whatever we selected as the morning levels of cytokines, the measurements were not conducted at identical times or with the same sample, biasing the possible association.(5) Although we tried to exclude some comorbidities that can influence the results, it is impossible to be sure that they were not influenced by others which we did not refer to.

Conclusions
This systematic review and meta-analysis of 102 articles provides substantial evidence that levels of proinflammatory cytokines TNF-α, IL-8, and IL-1β in adults and TNF-α, IL-8, and IFN-γ in children with OSA are higher than in controls.This study also identifies

Figure 1 .
Figure 1.Flowchart of the study selection.

Figure 1 .
Figure 1.Flowchart of the study selection.

Figure 2 .
Figure 2. Forest plot analysis of comparison of serum/plasma levels of TNF-α in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the

Figure 2 .
Figure 2. Forest plot analysis of comparison of serum/plasma levels of TNF-α in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 3 .
Figure 3. Forest plot analysis of comparison of serum/plasma levels of IL-8 in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 4 .
Figure 4. Forest plot analysis of comparison of serum/plasma levels of IL-1β in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 3 .
Figure 3. Forest plot analysis of comparison of serum/plasma levels of IL-8 in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 3 .
Figure 3. Forest plot analysis of comparison of serum/plasma levels of IL-8 in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 4 .
Figure 4. Forest plot analysis of comparison of serum/plasma levels of IL-1β in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 4 .
Figure 4. Forest plot analysis of comparison of serum/plasma levels of IL-1β in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 5 .
Figure 5. Forest plot analysis of comparison of serum/plasma levels of IFN-γ in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 6 .
Figure 6.Forest plot analysis of comparison of serum/plasma levels of TNF-α in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 5 .
Figure 5. Forest plot analysis of comparison of serum/plasma levels of IFN-γ in adults with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 6 .
Figure 6.Forest plot analysis of comparison of serum/plasma levels of TNF-α in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 6 .
Figure 6.Forest plot analysis of comparison of serum/plasma levels of TNF-α in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 7 .
Figure 7. Forest plot analysis of comparison of serum/plasma levels of IL-8 in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 8 .
Figure 8. Forest plot analysis of comparison of serum/plasma levels of IL-1β in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 9 .
Figure 9. Forest plot analysis of comparison of serum/plasma levels of IFN-γ in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 7 .
Figure 7. Forest plot analysis of comparison of serum/plasma levels of IL-8 in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 7 .
Figure 7. Forest plot analysis of comparison of serum/plasma levels of IL-8 in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 8 .
Figure 8. Forest plot analysis of comparison of serum/plasma levels of IL-1β in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 9 .
Figure 9. Forest plot analysis of comparison of serum/plasma levels of IFN-γ in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 8 .
Figure 8. Forest plot analysis of comparison of serum/plasma levels of IL-1β in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 7 .
Figure 7. Forest plot analysis of comparison of serum/plasma levels of IL-8 in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 8 .
Figure 8. Forest plot analysis of comparison of serum/plasma levels of IL-1β in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 9 .
Figure 9. Forest plot analysis of comparison of serum/plasma levels of IFN-γ in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Figure 9 .
Figure 9. Forest plot analysis of comparison of serum/plasma levels of IFN-γ in children with OSA compared to controls.Each square represents the result of an individual study.The size of the square indicates the weight of the study in the meta-analysis.Horizontal lines represent the confidence intervals of the individual studies.The longer the lines, the wider the confidence interval, indicating less reliability of the study results.The diamond represents the pooled result from all the studies.The width of the diamond indicates the 95% confidence interval for the pooled result.

Table 1 .
Characteristics of the articles including Adults.

Table 2 .
Characteristics of articles including children.

Table 3 .
Cont.Bold number means statistically significant (p < 0.05).IL: interleukin.TNF-α: tumor necrosis factor-alpha.AHI: apnea-hypopnea index.BMI: body mass index.In table, mean BMI and mean age included both groups (cases and controls) together.N: number of studies.