摘要目的:探讨MSM儿童期性虐待（CSA）与HIV感染风险的关联，并探索亲密性伴暴力（IPV）在两者关系中的中介作用，为针对性干预策略的制定提供参考依据。方法:采用横断面研究设计，2023年7月至2024年1月在武汉市性少数人群社会组织的协助下，采用方便抽样和同伴推动抽样法招募MSM共1 321人。采用问卷星手机版设计的调查问卷收集研究对象的社会人口学特征、CSA、IPV及HIV高危行为等信息。采用综合风险评估工具计算HIV感染风险评分，运用多元线性回归分析CSA与IPV对高危行为的影响，并采用Bootstrap法检验IPV的中介效应，按IPV类型、不同HIV高危行为进行亚组分析。结果:在MSM研究对象1 054人中，经历过CSA和IPV 的自我报告率分别为22.8%（240/1 054）和14.8%（158/1 054）。在HIV高危行为中，最近6个月有同性性伴数为2~5个、最近6个月发生无保护肛交、并行多性伴的自我报告率分别为59.1%（623/1 054）、49.2%（519/1 054）、26.3%（277/1 054）。多元线性回归模型分析结果显示，经历过CSA和IPV的MSM有更高的HIV感染风险评分（ β=0.437，95% CI：0.051~0.823； β=0.465，95% CI：0.008~0.920）。IPV类型的亚组分析发现，遭受型IPV在CSA与HIV感染风险评分关系中发挥部分中介效应（ β=0.045，95% CI：0.002~0.110），中介效应占总效应的10.3%。HIV高危行为的亚组分析发现，在CSA和无保护肛交关系中，综合型、遭受型和实施型IPV（ β=0.015，95% CI：0.002~0.030； β=0.014，95% CI：0.001~0.020； β=0.017，95% CI：0.002~0.030）均发挥完全中介效应。在CSA与并行多性伴的关系中，综合型和实施型IPV（ β=0.009，95% CI：0.000~0.020； β=0.011，95% CI：0.001~0.020）发挥完全中介效应。 结论:经历过CSA的MSM直接增加了HIV感染风险，也可通过IPV间接增加HIV感染风险，并且IPV的中介效应呈现类型与行为的特异性。

abstractsObjective:To explore the association between childhood sexual abuse (CSA) and HIV infection risk among men who have sex with men (MSM), and to examine the mediating role of intimate partner violence (IPV) in the relationship, thereby providing a reference basis for targeted interventions.Methods:A cross-sectional study design was adopted. From July 2023 to January 2024, with the assistance of social organizations for sexual minorities in Wuhan, a total of 1 321 MSM were recruited by convenient sampling and peer-driven sampling methods. The questionnaire designed by the mobile version of Wenjuanxing was used to collect information such as the sociodemographic characteristics, CSA, IPV and HIV high-risk behaviors of subjects. The HIV infection risk score was calculated using the comprehensive risk assessment tool. The effects of CSA and IPV on high-risk behaviors were analyzed by multiple linear regression, and the mediating effect of IPV was tested by the Bootstrap method. Subgroup analyses were conducted according to IPV types and different HIV high-risk behaviors.Results:Among the 1 054 subjects of MSM, the self-reporting rates of those who had experienced CSA and IPV were 22.8% (240/1 054) and 14.8% (158/1 054), respectively. Among high-risk behaviors of HIV, the self-reporting rates of having 2-5 homosexual sexual partners in the last 6 months, unprotected anal sex in the last 6 months, and having multiple sexual partners were 59.1% (623/1 054), 49.2% (519/1 054), and 26.3% (277/1 054), respectively.The multiple linear regression model analysis showed that MSM who had experienced CSA and IPV had a higher HIV infection risk score ( β=0.437, 95% CI: 0.051-0.823; β=0.465, 95% CI: 0.008-0.920).Subgroup analysis of IPV types revealed that suffering IPV played a partial mediating role in the relationship between CSA and HIV infection risk score ( β=0.045, 95% CI: 0.002-0.110), with the mediating effect accounting for 10.3% of the total effect. Subgroup analysis of high-risk behaviors for HIV revealed that in CSA and unprotected anal sex relationships, comprehensive, suffered and administered IPV ( β=0.015, 95% CI: 0.002-0.030; β=0.014, 95% CI: 0.001-0.020; β=0.017, 95% CI: 0.002-0.030) all exerted a complete mediating effect. In the relationship between CSA and parallel multiple sexual partners, the comprehensive and implemented IPV ( β=0.009, 95% CI: 0.000-0.020; β=0.011, 95% CI: 0.001-0.020) exert a complete mediating effect. Conclusions:MSM who have experienced CSA directly increase the risk of HIV infection, and can also indirectly increase the risk of HIV infection through IPV. Moreover, the mediating effect of IPV showed specificity in terms of type and behavior.