甲状腺功能障碍与前列腺癌的因果关联：两样本孟德尔随机化研究

Causal relationship between thyroid dysfunction and prostate cancer： a two-sample Mendelian randomization study

摘要:

目的 采用两样本的孟德尔随机化（two-sample Mendelian randomization，TSMR）的分析方法，探讨甲状腺功能障碍与前列腺癌之间的因果关联。方法 利用全基因组关联研究（genome-wide association studies，GWAS）的数据进行分析，暴露因素为甲状腺功能障碍，包括甲状腺功能亢进（甲亢）、甲状腺功能减退（甲减），结局变量为前列腺癌。采用逆方差加权分析（inverse-variance weighted，IVW）法、加权中位数（weighted median，WM）、MR Egger法进行孟德尔随机化分析，并进行异质性检验、敏感性分析、多效性分析。结果 甲亢与前列腺癌存在负向的因果关系（IVW：OR=0.015，95%CI=0~0.436），先天性甲减与前列腺癌存在正向的因果关系（IVW：OR=1.088，95%CI=1.003~1.180），其他类型的甲减并未发现与前列腺癌存在因果关系。且所有结果均不存在多效性，留一法也提示所得结果稳健。结论 甲状腺功能障碍与前列腺癌存在因果关系，低水平的甲状腺激素可能与前列腺癌风险增加有关，但仍需更大的GWAS数据通过进一步的孟德尔随机化研究来验证这种关系。

Abstract:

Objective To investigate the causal relationship between thyroid dysfunction and prostate cancer by using the two-sample Mendelian randomization method.Methods Relevant genome-wide association study（GWAS） data were collected. The exposure factor was thyroid dysfunction（including hyperthyroidism，hypothyroidism），and the outcome variable was prostate cancer. We performed Mendelian randomization analysis using the inverse-variance weighted method（IVW），weighted median，and Mendelian randomization-Egger method. The heterogeneity test，sensitivity analysis，and pleiotropic analysis were also carried out.Results There was a negative causal relationship between hyperthyroidism and prostate cancer（IVW：OR=0.015，95%CI=0~0.436） and a positive causal relationship between congenital hypothyroidism and prostate cancer（IVW：OR=1.088，95%CI=1.003~1.180）. No causal relationship was found between the other types of hypothyroidism and prostate cancer. No results had pleiotropy. Leave-one-out cross-validation suggested that the results were robust.Conclusion Thyroid dysfunction is causally linked to prostate cancer，and low levels of thyroid hormones may be related to an increased risk of prostate cancer，which need further verification through Mendelian randomization studies using larger-size GWAS data.