Introduction Globally, substantial heterogeneity in the prevalence of HIV and other sexually transmitted infections (STIs) among most at-risk populations (MARPs) has been demonstrated. Examining factors related to heterogeneity can inform targeted programming. In Winnipeg, Canada, particularly high risk for HIV and hepatitis C (HCV) has been observed in injection drug users (IDUs) with a history of solvent use (S-IDUs). However, comparisons to other MARPs have been limited. Thus this study examined the association between HIV/STIs and S-IDUs in comparison to IDUs and other MARPs.

Methods Data were from a 2008 to 2009 cross-sectional study of Winnipeg MARPs (IDUs, sex work- and street-involved individuals); subjects were recruited through respondent-driven sampling (RDS) methods. Adjusted ORs (AORs) from multivariable logistic regression models were estimated, examining the risk of HIV, HCV and HIV/HCV co-infection, and corrected for RDS-chain clustering using generalised estimating equations.

Results Total sample was 499, of which 13% recently injected drugs (ie, last 6 months), 5% recently inhaled solvents, 6% were recent S-IDUs, and 76% did not inject drugs or inhale solvents. HIV and HCV prevalence among recent S-IDU was 21% and 79%, respectively; HIV/HCV co-infection was 18%. In multivariable models, S-IDUs were at highest risk of HIV (AOR: 3.6, 95% CI: 1.6% to 7.9%; p<0.001), HCV (AOR: 19.3, 95% CI: 6.8% to 58.3%; p<0.001) and HIV/HCV co-infection (AOR: 6.0, 95% CI:2.5% to 14.7%; p<0.001). Comparatively, AORs for IDU-only were 3.3 (95% CI: 1.3% to 7.6%), 3.8 (95% CI: 2.3% to 7.5%) and 4.9 (95% CI: 2.1% to 14.7%). Among lifetime S-IDUs, elevated risk for HIV (AOR: 7.4, 95% CI: 2.3% to 26.2%) and HCV (AOR: 22.7, 95% CI: 11.0% to 47.0%) was observed, but not for HIV/HCV co-infection.

Conclusions Solvent use occurs among the most marginalised of MARPs, representing unique and complicated drug use trajectories. As the HIV epidemic in Canada becomes increasingly complex, examination of outlier populations such as S-IDU can inform public health by elucidating important pathways by which structural, environmental and individual factors interact to create the highest risk for HIV/STIs and other bloodborne pathogens.