Cardiovascular diseases are the major cause of death and disability worldwide. According to the World Health Organization report in 2011, more than 17 million people died of cardiovascular diseases. Coronary artery disease (CAD) is characterized by occlusive epicardial coronary artery stenosis. Ischemic stroke (IS) is a major kind of stroke, which causes a high death rate and adult disability in the world. According to the report of Liu, the annual stroke mortality rate was 120–180 per 100,000 in China [1
]. In the United States, stroke ranks as the third leading cause of death [2
]. IS has exceeded heart diseases to become the most frequent cause of death. IS and CAD are principal clinical manifestations of atherosclerosis, and have caused a huge burden for society. Previous studies demonstrated that single nucleotide polymorphisms (SNPs) were associated with the risks of CAD and IS [3
MicroRNAs are small, non-coding RNAs which regulate the gene expression in post-transcriptional levels. The regulation effects of microRNAs are obtained by binding to the 3ʹ-UTR of target mRNAs, and lead to degradation or translation repression of target genes. Previous reports predicted that 1/3 of human genes were regulated by microRNAs [5
]. Thus, the microRNAs interfere with many physiologic and pathological processes. The changes in the sequences of microRNAs, such as single nucleotide polymorphisms (SNPs), may result in diseases.
miR-146a is a microRNA located at the human chromosome 5q33. Previous studies have demonstrated that miR-146a participates in inflammatory processes, thus interferes with the pathology of cardiovascular diseases [6
]. A SNP has been found to exist at the precursor miRNA-146a, which mutates G:U to C:U and results in a low production of mature miR-146a [7
]. Studies have reported the relationship between miR-146a polymorphism and susceptibility of coronary artery disease and ischemic stroke. However, the results are somehow inconsistent. For example, some studies demonstrated that miR-146a polymorphism was associated with the risk of CAD and IS [8
], while others thought it was not [13
Thus it is necessary to make a precise and comprehensive estimation of the association between miR-146a and the risks of CAD and IS. In the present study, eight studies, including 3138 cases and 3097 controls, were included in the meta-analysis.
The main findings of our meta-analysis show that rs2910164 is associated with the risk of CAD significantly in allelic model (OR = 0.86), homozygous model (OR = 0.70), heterozygous model (OR = 0.80) and dominant model (OR = 0.76). The subjects carrying GG genotype, GG + GC genotype or G allele are at lower risks of CAD. For the susceptibility of IS, there are no significant associations between rs2910164 and total studies. However, in subgroup analysis by sample size and ethnicity, the GG, GG + GC and G allele of rs2910164 are found to be associated with higher risks of IS in large size group and in Koreans, under homozygous and dominant model.
Although a very recent meta-analysis was performed to reveal the relationship of rs2910164 and cardio-cerebrovascular diseases, He et al.
], they included fewer studies on CAD and IS than those in our meta-analysis. In our meta-analysis, we included two more studies, with 737 cases and 731 controls, for CAD and IS analysis. We further conducted a more comprehensive subgroup analysis for IS by TOAST, gender, smoking, hypertension, diabetes mellitus and hyperlipidemia. Furthermore, we found some surprisingly different results between our meta-analysis and He et al.
’s in CAD. Although we found a similar lower CAD risks for G allele vs.
C allele, our meta-analysis also demonstrated a lower risks of CAD for GG/GC and GG genotype vs.
CC carriers (Figure 2
and Table 2
). However, He et al
.’s meta-analysis stated a higher risk for GG carriers vs.
In the sensitivity analysis, no significant changes were found after omitting each study one at a time, indicating the relative stability and credibility of the results of our meta-analysis.
We conducted a subgroup analysis by sample size and ethnicity, and observed that rs2910164 was associated with the risks of CAD and IS in the large sample group. However, only one study was included in the large size subgroup in CAD, and two studies were included in IS. We also found a higher risk of IS in Korean for G allele, GG/GC and GG genotypes of miR-146a, and once again, only one study is included in this subgroup. Thus, further studies are needed to confirm these results.
Since no associations were found between rs2910164 and IS, we performed subgroup analysis by TOAST, gender, smoking, hypertension, diabetes mellitus and hyperlipidemia (Table 3
). No significant relations were found in all the analyses. But some studies lack sufficient data for the subgroup analysis. For example, the study of Huang et al
] lacks TOAST data, while Jeon et al
. and Zhu et al
. lack gender, smoking, hypertension, hyperlipidemia and diabetes mellitus data [12
]. Interestingly, in the study of Zhu et al
], no significant correlations were found between rs2910164 and overall subjects, though when they divided the patients to LAA and SVD according to TOAST typing, significantly higher frequencies in LAA-caused IS were found in the CC genotype and C allele. However, the results of Zhu et al
], focusing on LAA-caused IS was contradictory to the study of Jeon et al
]. This may be the result of different target genes, geographical ethnic groups, sample size, etc.
Further studies are still needed to resolve this discrepancy.
CAD and IS are both manifestations of atherosclerosis. Most studies included in the present meta-analysis found dyslipidemia in the case group. Previous studies have indicated atherosclerosis as an inflammatory process. Thus, anti-inflammatory treatments decrease the risk of atherosclerosis. miR-146a regulates the NF-κB-induced inflammatory process by targeting interleukin-1 receptor-associated kinase 1 (IRAK-1) and TNF receptor-associated factor 6 (TRAF-6). IRAK-1 and TRAF-6 are upstream regulators of NF-κB activation, and major signal transducers of the Toll-like receptor (TLR) system [9
]. The SNP in the pre-miR-146a changed the G to C allele, thus influence the mature miR-146a production, and subsequently influence the inflammatory process of atherosclerosis. Interestingly, the GG, GG + GC genotype and G allele are associated with lower susceptibility of CAD in overall subjects and large sample size group. On the contrary, GG, GG + GC genotype and G allele are related to higher risk of IS in large sample size group. This could be explained by the target gene selection of miR-146a and the pathological differences of these two diseases [10
]. Furthermore, the endothelial dysfunction induced by the inflammatory process is considered to be the first step of atherosclerosis. Further investigations are needed to reveal whether or not rs2910164 affects endothelial mature miR-146a production and the effects of this variation might have on endothelial functions.
The results of the present meta-analysis should be interpreted carefully because of the following limitations. Firstly, the number of patients was relatively small, and may influence the outcomes. After a very comprehensive literature search from several different databases, only a total of eight studies were included in the present meta-analysis. Among them, four are related to CAD (1265 cases and 1241 controls) and four are related to IS (1873 cases and 1856 controls). The overall OR only indicated the associations between rs2910164 and CAD. However, no associations were found between rs2910164 and IS risk. The second limitation is that the clinicopathological characteristics or disease subtypes are limited in most of these studies. Although we conducted subgroup analysis by clinicopathological characteristics in IS, only half of the included studies provide the necessary data. And for CAD, no such characteristics were available. Thirdly, CAD and IS are both multi-factorial diseases influenced by both genetic and environmental factors. The gene-gene and gene-environment interactions may play important roles in the function of rs2910164, but most studies lack information about environmental exposure and multiple SNPs in miRNA-encoding genes. The fourth limitation lies in the ethnicity of the subjects. Most of the patients were Asians in the present study and this limited the general application of the results to other populations.
In conclusion, the current meta-analysis suggests a decreased risk of CAD for GG, GG + GC genotype and G allele of rs2910164, while rs2910164 is not associated with the risk of IS. Thus rs2910164 might be recommended as a predictor for susceptibility of CAD, but not IS. However, the results of this meta-analysis should be interpreted with caution because of the heterogeneity among study designs. Further study is needed to evaluate the association of rs2910164 and these two diseases, especially in a large sample size, in Caucasians, and with clinicopathological characteristics.