Introduction:
Identifying patterns of human immunodeficiency viruses (HIV) transmission and factors affecting it can help to prevent and control the disease. The present study aimed to identify the latent classes of HIV risk factors in Iran.

Material and methods:
This cross-sectional study was conducted among 32,168 HIV patients. We fitted latent class analysis by considering 11 indicators. Models with 2 to 5 classes were fitted, and the best-fitted model was chosen based on Acaik information critical (AIC), Bayesian information criterion (BIC), entropy, and interpretability of the results. Additionally, multinomial logistic regression model was applied to assess relationships between the covariates and latent class membership.

Results:
We identified 3 latent classes, including low-risk (41.35%), high-risk (50.28%), and moderate-risk (8.37%) classes. The moderate-risk class was most likely prison history (69.09%) and addiction (67.37%). The high-risk class included addiction (99.68%), injection drugs users (IDUs) (99.06%), prison history (86.27%), and sharing needles (79.64%). Older age and being single increased the odds of membership in high-risk class (OR: 1.021, 95% CI: 1.017-1.024%; OR: 1.34, 95% CI: 1.20-1.50%), respectively, while older age increased the odds of membership in classes of moderate-risk (OR: 1.028, 95% CI: 1.023-1.033%) compared with low-risk group (p < 0.05). The odds of membership in classes of high-risk (OR: 0.011, 95% CI: 0.010-0.013%) and moderate-risk (OR: 0.113, 95% CI: 0.098-0.131%) were significantly lower in women than in men compared with low-risk group (p < 0.05).

Conclusions:
Injection drugs users, drug addiction, sharing needles, and prison history are related to HIV infection in Iran. Moreover, older age, female gender, low education, and being single increase the odds of membership in high-risk classes. These findings highlight the need for preventive interventions and harm reduction programs for people at risk.