Predicting factors of sexual behaviours at risk of HIV infection among people living with disabilities: a generalized structural equation model.

Many studies have shown the HIV vulnerability of people with disabilities (PWD). This vulnerability is supported by risky sexual behaviours, which need to be well described. The objective of this study is to identify predictors of sexual behaviours at risk of HIV infection among PWD in Burkina Faso. We conducted a population-based cross-sectional study on disability and HIV in four regions in Burkina Faso. PWD were selected through a probabilistic method. The Washington Group Short questionnaire was used to identify PWD. Five sexual risk behaviours were considered in this analysis (having multiple sex partners, starting sex at an early age, having had sex with a nonregular partner, sex in return for payment and condomless sex). A composite variable was created and called "risky sexual behaviour (RSB)", which was defined in three categories (low risk, moderate risk and high risk) based on the co-occurrence of the individual sexual risk behaviours. Generalized structural equation was used to identify the predictors of RSB withincome possession as a mediator. A total of 973 PWD were included in this study. The proportion of PWD who were engaged in at least two risky sexual behaviours was 9.5% (95%CI: 6.5-13.8), those with one risky sexual behaviour represented 42.8% (95%CI: 39.4-46.3) and less than half ( 47.7% (95%CI: 43.3-52.1)) werenot engaged in risky sexual behaviour. The prevalence of HIV increased with the level of risky sexual behaviour. Age, education level and type of disability have a significant direct effect on RSB. Indeed, the oddof engaging in sexual risk behaviour was low among PWD who attended at least secondary school compared to those who were not schooled (adjusted odds ratio (aOR): 0.62 (95%CI: 0.41-0.92). The indirect effect of sex on RSB mediating by income possession was significant. Indeed, women with income have a low oddof engaging in RSB compared to men (aOR: 0.83 [95%CI: 0.71-0.98]). Our results confirm a certain ambivalence in the sexuality of people with disabilities. Hence, there is a need for a particular focus on risky sexual behaviour to enable the development of an effective HIV strategy in this group.