Background: Observational studies support the inflammation hypothesis in coronary heart disease (CHD). As a pleiotropic proinflammatory cytokine, Interleukin-18 (IL-18), has also been found to be associated with the risk of CHD. However, to our knowledge, the method of Mendelian Randomization has not been used to explore the causal effect of IL-18 on CHD.

Objective: To assess the causal effect of IL-18 on the risk of CHD.

Methods and Results: Genetic variant instruments for IL-18 were obtained from information of the CHS and InCHIANTI cohort, and consisted of the per-allele difference in mean IL-18 for 16 independent variants that reached genome-wide significance. The per-allele difference in log-odds of CHD for each of these variants was estimated from CARDIoGRAMplusC4D, a two-stage meta -analysis. Two-sample Mendelian Randomization (MR) was then performed. Various MR analyses were used, including weighted inverse-variance, MR-Egger regression, robust regression, and penalized regression. The OR of elevated IL-18 associated with CHD was only 0.005 (95%CI -0.105~0.095; P-value=0.927). Similar results were obtained with the use of MR-Egger regression, suggesting that directional pleiotropy was unlikely biasing these results (intercept -0.050, P-value=0.220). Moreover, results from the robust regression and penalized regression analyses also revealed essentially similar findings.

Conclusion: Our findings indicate that, by itself, IL-18 is unlikely to represent even a modest causal factor for CHD risk.

背景：观察性研究支持冠心病（CHD）中的炎症假说。作为多效促炎性细胞因子，白介素18（IL-18）也被发现与冠心病风险有关。然而，据我们所知，孟德尔随机化方法尚未用于探讨IL-18对冠心病的因果作用。

目的：评估IL-18对冠心病风险的因果关系。

方法和结果：从CHS和InCHIANTI队列的信息中获得了IL-18的遗传变异工具，该工具由16个独立变异的平均IL-18的等位基因差异组成，这些变异达到了全基因组范围。根据两个阶段的荟萃分析，CARDIoGRAMplusC4D估算了这些变体中每一个的CHD对数奇数的等位基因差异。然后进行两样本孟德尔随机化（MR）。使用了各种MR分析，包括加权反方差，MR-Egger回归，鲁棒回归和惩罚回归。与冠心病相关的IL-18升高的OR仅为0.005（95％CI -0.105〜0.095； P值= 0.927）。通过使用MR-Egger回归获得了类似的结果，表明方向多效性不太可能使这些结果产生偏差（截距-0.050，P值= 0.220）。

结论：我们的发现表明，IL-18本身就不可能代表冠心病风险的适度因果因素。