Association between the ESR1 and ESR2 polymorphisms and osteoporosis risk: An updated meta-analysis

BACKGROUND: Gene polymorphisms of estrogen receptor (ESR) 1 PvuII (rs2234693), XbaI (rs9340799), G2014A (rs2228480), ESR2 AluI (rs4986938), and RsaI (rs1256049) had been reported to be associated with the risk of osteoporosis. However, these conclusions were inconsistent, therefore, an updated meta-analysis was conducted to further explore these issues. OBJECTIVE: To evaluate the association between gene polymorphisms of ESR1 PvuII (rs2234693), XbaI (rs9340799), G2014A (rs2228480), ESR2 AluI (rs4986938), RsaI (rs1256049), and osteoporosis risk. MATERIALS AND METHODS: PubMed, Medline, Ovid, Embase, CNKI, and China Wanfang databases were searched. Association was assessed using odds ratio with 95% confidence interval. Moreover, the false-positive reporting probability, Bayesian false-finding probability, and Venetian criteria were used to assess the credibility of statistically significant associations. RESULTS: Overall, ESR1 PvuII (rs2234693) and XbaI (rs9340799) were associated with the risk of osteoporosis in Indians. Moreover, ESR1 G2014A (rs2228480) was associated with the decreased risk of osteoporosis in East Asians. Moreover, ESR2 Alul (rs4986938) was associated with the increased risk of osteoporosis in East Asians and Caucasians. There was a significant association between ESR2 Rsal (rs1256049) and osteoporosis risk in overall population. When only high-quality and Hardy–Weinberg equilibrium studies were included in the sensitivity analysis, all results did not change in the present study. When the credibility was evaluated applying false-positive reporting probability, Bayesian false-finding probability, and Venetian criteria, all significant associations were considered as false positive results. CONCLUSIONS: In summary, this study shows that all substantial associations between gene polymorphisms of ESR1 (PvuII, XbaI, and G2014A) and ESR 2 (AluI and RsaI) and osteoporosis risk are possibly false positive results instead of real associations or biological variables.