[Latent-class analysis of intimate partner violence and HIV high risk behaviors among college students in Zhuhai].

<b>Objective:</b> To explore the latent-classes of HIV high risk behaviors among college students, and the association between experiences of intimate partner violence (IPV) and HIV high risk behaviors, to provide evidence for reducing the HIV high risk behaviors among them. <b>Methods:</b> A cross-sectional study was conducted from October to December 2019 among university students from six higher education institutions in Zhuhai City, using a multi-stage cluster sampling method, with an estimated sample size of 1 318. The study included participants who self-reported being in a romantic relationship and having sexual experience within the past year. Data on sociodemographic characteristics, IPV experiences, and HIV high risk behaviors were collected. Latent-class analysis was performed on HIV high risk behaviors, and chi-squared tests and multivariable logistic regression were used to analyze the associations between IPV experiences and different latent classes of HIV high risk behaviors. <b>Results:</b> The effective response rate for the survey was 95.4% (12 235/12 821). 1 382 college students from Zhuhai were included as participants in the study, with 19.4% (268/1 382) self-reporting having experienced IPV. Latent-class analysis of HIV high risk behaviors classified the participants into three latent groups: low-risk group (78.1%, 1 079/1 382), multiple sexual partners/alcohol use before sex group (15.8%, 219/1 382), and high-risk group (6.1%, 84/1 382). Multivariable logistic regression analysis showed that students who had experienced psychological violence were more likely to be in the group that had multiple sexual partners/alcohol use before sex (a<i>OR</i>=2.51, 95%<i>CI</i>:1.48-4.27). Those who had experienced IPV (a<i>OR</i>=5.74, 95%<i>CI</i>:3.45-9.55), physical violence (a<i>OR</i>=9.26, 95%<i>CI:</i> 5.24-16.35), sexual violence (a<i>OR</i>=8.46, 95%<i>CI</i>:4.93-14.52), or psychological violence (a<i>OR</i>=15.99, 95%<i>CI</i>:8.64-29.57) were more likely to be in the high-risk group. Students who experienced two (a<i>OR</i>=9.37, 95%<i>CI</i>:3.55-24.71) or three types of IPV (a<i>OR</i>=50.09, 95%<i>CI</i>: 21.06-119.14) were more likely to be in the high-risk group compared to those with no IPV experiences. <b>Conclusions:</b> HIV high risk behaviors among college students in Zhuhai exhibited heterogeneity across different latent groups, and these groups have different associations with IPV experiences. Universities should tailor targeted HIV/AIDS education and prevention strategies based on the characteristics of each latent group to reduce HIV high risk behaviors among college students.