目的·采用双样本孟德尔随机化（Mendelian randomization，MR）方法验证孤独、主观幸福感与衰弱指数的关联是否反映了因果关系。方法·基于已发表的全基因组关联研究（genome-wide association studies，GWAS）的摘要数据，利用MR方法分析孤独感、主观幸福感与衰弱指数（作为衰弱的替代指标）的因果关系。孤独感数据来源于包含445 024名欧洲血统参与者的研究，包括80 134名欧洲血统病例和364 890名欧洲血统对照；主观幸福感数据来自298 420名欧洲血统参与者；衰弱的汇总数据来自英国生物银行和瑞典TwinGene GWAS的meta分析，包括175 226名欧洲血统的参与者。使用逆方差加权（inverse variance weighted，IVW）方法作为主要的MR分析方法确定估计值，加权中位数（weighted median estimator，WME）作为补充方法。采用Cochran′s Q检验检查IVW估计值中异质性，并使用MR Egger截距检验指示多效性的存在。存在异质性时使用异常值全局检测MR-PRESSO识别离群值，去除离群值后，重复MR分析。进行留一分析，以评估总体估计值是否由单一的单核苷酸多态性（single nucleotide polymorphism，SNP）驱动。在暴露和结局数据的效应方向校正过程中，剔除等位基因方向难以确定的回文SNP以确保数据的一致性和分析的可靠性。结果·通过IVW证实孤独感与衰弱指数呈显著正相关，OR为3.87（95% CI 2.33~6.46，P<0.001）；WME进一步验证了这一结果，OR为2.81（95% CI 1.49~5.29，P<0.001）。主观幸福感与衰弱指数呈显著负相关，OR为0.80（95% CI 0.69~0.94，P=0.005）；WME结果趋势一致，OR为0.86（95% CI 0.74~1.01，P=0.065）。未发现异质性（Cochran′s Q检验，P=0.054和P=0.074）或水平多效性（MR-Egger截距检验，P=0.470和P=0.260）。结论·通过MR方法强化了孤独感增加衰弱风险、主观幸福感对衰弱具有保护作用的因果证据。

Objective ·To employ a two-sample Mendelian randomization (MR) design to examine whether the associations between loneliness, subjective well-being, and the frailty index reflect causal relationships. Methods ·Summary data from published genome-wide association studies (GWAS) were utilized, and an MR approach was employed to analyze the causal relationships between loneliness, subjective well-being, and the frailty index (as a surrogate measure of frailty). Data on loneliness were derived from a study comprising 445 024 individuals of European ancestry, including 80 134 cases and 364 890 controls. Data for subjective well-being were obtained from 298 420 participants of European ancestry. Summary statistics for frailty were sourced from a meta-analysis of GWAS conducted within the UK Biobank and the Swedish TwinGene cohort, which included 175 226 individuals of European ancestry. The inverse variance weighted (IVW) method served as the primary MR analytical approach to derive causal estimates, with the weighted median estimator (WME) used as a supplementary approach. Cochran′s Q test was applied to assess heterogeneity in the IVW estimates. The presence of horizontal pleiotropy was evaluated using the MR-Egger intercept test. When significant heterogeneity was detected, the MR-PRESSO global test was utilized to identify and subsequently remove outlier variants. MR analyses were then repeated. A Leave-One-Out sensitivity analysis was conducted to evaluate whether the overall estimates were unduly influenced by any single nucleotide polymorphism (SNP). To ensure data consistency and analytical reliability during effect allele harmonization between exposure and outcome datasets, palindromic SNPs (for which the effect allele direction could not be definitively determined) were excluded. Results ·The IVW method revealed a significant positive association between loneliness and the frailty index, with an OR of 3.87 (95% CI 2.33‒6.46, P<0.001). This result was further confirmed by the WME, yielding an OR of 2.81 (95% CI 1.49‒5.29, P<0.001). Subjective well-being showed a significant negative association with the frailty index (OR=0.80, 95% CI 0.69‒0.94, P=0.005). The WME demonstrated a consistent direction of effect, though with borderline significance (OR=0.86, 95% CI 0.74‒1.01, P=0.065). The analysis revealed no significant heterogeneity (Cochran′s Q test, P=0.054 and P=0.074) or horizontal pleiotropy (MR-Egger intercept test, P=0.470 and P=0.260). Conclusion ·This study, supported by the MR methodology, indicates that loneliness is associated with an increased risk of frailty, while subjective well-being serves as a protective factor against frailty.