目的 评估锌、镁等矿物质水平与孤独症谱系障碍（autism spectrum disorder, ASD）的因果关联。 方法 采用两样本孟德尔随机化（Mendelian randomization, MR）方法，基于欧洲人群全基因组关联研究汇总数据（18 382例ASD患儿和27 969例对照）以及UK Biobank铁、钙、镁水平遗传数据及澳英联合队列的锌、硒遗传数据，采用MR分析框架，筛选351个遗传工具变量，使用逆方差加权法为主要分析方法进行因果推断，并采用Cochran's Q检验、MR-PRESSO全局检验等方法进行敏感性分析。 结果 循环锌、镁、钙、硒、铁与ASD的因果效应均未显示统计学意义（P>0.05），其逆方差加权法分析比值比及95%置信区间分别为0.934（0.869~1.003）、1.315（0.971~1.850）、1.055（0.960~1.159）、1.015（0.953~1.080）和0.946（0.687~1.303）。敏感性分析表明循环钙与ASD的因果关联存在显著异质性（P=0.006），但MR-PRESSO校正后效应值稳定（P=0.487），其余矿物质的因果效应估计具有较好的稳健性。 结论 该研究未发现显著证据支持循环锌、镁、钙、硒、铁水平与ASD风险之间存在因果关联，为ASD的病因学研究和精准营养干预提供了重要线索。

Objective To evaluate the causal association between circulating levels of zinc, magnesium, and other minerals and autism spectrum disorder (ASD). Methods A two-sample Mendelian randomization (MR) analysis was performed using summary statistics from large-scale genome-wide association studies of European populations, including 18 382 ASD cases and 27 969 controls. Genetic data for iron, calcium, and magnesium were obtained from the UK Biobank, and data for zinc and selenium were sourced from an Australian-British cohort. A total of 351 genetic instrumental variables were selected. Causal inference was performed using inverse-variance weighting as the primary analysis method. Sensitivity analyses were performed by Cochran's Q test and MR-PRESSO global test to assess the robustness of the findings. Results No statistically significant causal effect was observed for circulating zinc, magnesium, calcium, selenium, or iron levels on ASD risk (all P>0.05). The odds ratios and 95% confidence intervals from the inverse-variance weighting analysis were 0.934 (0.869-1.003) for zinc, 1.315 (0.971-1.850) for magnesium, 1.055 (0.960-1.159) for calcium, 1.015 (0.953-1.080) for selenium, and 0.946 (0.687-1.303) for iron. Sensitivity analysis revealed significant heterogeneity in the causal association between circulating calcium and ASD (P=0.006), while the effect estimate remained stable after MR-PRESSO correction (P=0.487). The causal effect estimates for the remaining minerals demonstrated good robustness. Conclusions This study did not find significant evidence supporting a causal association between circulating zinc, magnesium, calcium, selenium, or iron levels and ASD risk, providing important clues for the etiology of ASD and precision nutritional interventions.