The present meta-analysis demonstrated that IL-6 rs1800796 polymorphism was not associated with risk of lung cancer under all genetic models. In the subgroup analysis, the non-significant association remained in either never-smokers or smokers.
A previous meta-analysis reported that there was no significant association between IL-6 level and lung cancer risk . However, higher level of IL-6 is suggested to be associated with risk of coronary heart disease  and type 2 diabetes . In regards to the functional polymorphism rs1800796 in the IL-6 gene, it was only found to be associated with risk of type 2 diabetes , but not with coronary heart disease .
It is consistent that rs1800795 polymorphism was not associated nearly all types of cancers, including colorectal , prostate , gastric  and breast cancers . But rs1800796 polymorphism was found to be associated with prostate  cancer but not with gastric cancer . Based on 2691 cases with lung cancer and 3067 normal controls, we did find significant association between rs1800796 polymorphism and lung cancer.
Heterogeneity may potentially affect the interpretation of the results. Heterogeneity may be attributed to the potential confounding resulted from publication time, sample sizes, measurement errors, or the interaction with other risk factors. In our study, there was significant between-study heterogeneity for association between IL-6 rs1800796 polymorphism and risk of lung cancer under all genetic models. However, meta-regression analysis in consideration of the potential confounders did not address the heterogeneity.
There are several limitations in the present study. First, there might be effects of gene–gene and gene–environment interactions [27–29] but we can not address this because the individual studies did not provide the related data. Second, the sample size in the subgroups was limited and the results should be interpreted with caution. Third, we only assessed one polymorphism in the IL-6 gene, therefore, we can not rule out the possibility that other polymorphisms or haplotypes in this gene might be implicated in the development of lung cancer. Fourth, all the six papers we selected were from East Asia. Thus, the conclusion should not be generalized to other ethnic populations.