Basic characteristics of included studies.

Although numerous studies have suggested the association between TNF-α-308G/A polymorphism and susceptibility to obstructive sleep apnea (OSA), the results remained controversial and ambiguous. We performed the present meta-analysis to derive a more precise estimation.The PubMed, Embase, Cochrane library, Web of Science, China National Knowledge Infrastructure, Wanfang databases, and Weipu databases (until January 8, 2022) were accessed to retrieve relevant articles. Pooled odds ratios (ORs) with 95% confidence interval (95% CI) were calculated using the STATA statistical software.Totally, fourteen studies involving 2595 cases and 2579 controls were enrolled in this meta-analysis. Pooled results demonstrated significant association between TNF-α-308G/A polymorphism and OSA risk for the overall population(allele model:OR = 1.87 [1.47, 2.38] (n = 14), dominant model: OR = 1.88[1.48, 2.39] (n = 14), recessive model:OR = 2.83 [2.00, 4.00] (n = 11), homozygous model:OR = 3.30 [2.32, 4.68] (n = 11), and heterozygous model:OR = 1.67 [1.36, 2.06] (n = 14); P<0.001, respectively).Subgroup analysis showed that in both Caucasians and Asians, the A allele conferred increased risk to OSA compared to the G allele (Caucasians: OR = 1.40[1.03, 1.90] (n = 5), P = 0.033, Asians: OR = 2.30 [1.62, 3.26] (n = 9), P< 0.001). In subgroup analysis restricted to hospital-based individuals, significant association between TNF-α-308G/A polymorphism and OSA risk was identified under each genetic model. Whereas, in population-based individuals, increased risk of OSA were only found in homozygous model (OR = 2.19[1.23, 3.90] (n = 3), P = 0.008) and recessive model (OR = 1.77 [1.00, 3.13] (n = 3), P = 0.048). There was a substantial between-study heterogeneity (I2 = 69.10%) across studies which was explained by source of control participants (P = 0.036) by meta-regression. The results of leave-one-out meta-analysis and publication bias suggested the reliability and stability of our results.This meta-analysis suggested that TNF-α-308A allele may be a risk factor for the development of OSA. However, large scale,multi-center and well-designed case-control studies are needed in the future.