In the present study, 766 (75 %) youth reported recently engaging in unprotected sex during the study period and the majority of study observations included a report of recent unprotected sex. Female gender, Caucasian ancestry, substance use, monogamous relationships, and having multiple concurrent sex partners, were independently and positively associated with unprotected sex. The high prevalence of unprotected sex in this study aligns with previous findings that up to 25 % of street-youth have never used condoms and 56 % did not use condoms the last time they had sex under the influence of substances [10, 21]; this contrasts with a much higher proportion of condom use at last intercourse among the general youth population aged 20-24 in 2009/2010 (63 %) [8]. It is unclear, however, if our study outcome of “any unprotected sex in the recent six months” is comparable to “unprotected sex at last sexual intercourse”.

A number of different drugs were positively and significantly associated with unprotected sex in our study. Experimentation with alcohol at an early age is common among young people [22], and this study found that youth who reported alcohol use were more likely to report having unprotected sex. Alcohol is known to lower inhibitions which increases the likelihood of engaging in sexual activities that one might not normally partake in, such as sexual encounters with strangers, anal intercourse, and sex without a condom [23–25].

The null findings for binge drug use and injection drug use in the current analysis indicate that youth in our sample who engage in unprotected sex are not more likely to engage in especially risky drug use patterns. However, crystal methamphetamine and cocaine use were significantly associated with engaging in unprotected sex in our analysis. This is consistent with previous research findings that stimulant drug use heightens sexual arousal and lowers inhibitions, resulting in a higher likelihood of engaging in risky sexual behavior [26, 27]. The link between stimulant drug use, increased sexual arousal and reduced inhibitions, resulting in lower condom use, is particularly well documented in the context of sexual health among men who have sex with men [28]. The effect of stimulant use has also been found to increase risk of STI transmission among adults who use illicit drugs, in part by facilitating longer periods of sexual activity which can lead to increased risk of condom breakage [27, 29].

Our results provide further evidence that reducing stimulant drug use may prevent high levels of unprotected sex among this population. It is therefore of concern that vulnerable youth report high rates of difficulty accessing addiction treatment [30, 31]. Sustained efforts to improve engagement and retention in addiction treatment are warranted and can be expected to have positive health benefits beyond reductions in substance use [32, 33]. For youth who are unable or unwilling to reduce engagement in stimulant drug use, alternative interventions are needed. There is some evidence to suggest that low-threshold services such as supervised injection facilities [34] and needle exchange programs [35] may increase condom use; however, more studies to assess whether these secondary benefits would be realized with street-involved youth are needed. In addition, research indicates that the risks of HIV transmission through sexual intercourse can be reduced through expanded HIV testing and treatment [36]. Consequently, STI and HIV testing for vulnerable youth who use stimulants should be a public health priority and integrated into all healthcare services.

It is noteworthy that ecstasy and marijuana use were each also positively and significantly associated with unprotected sex in this study. Ecstasy is known to induce feelings of euphoria, friendliness, and enhanced sensuality [37], and previous research has linked ecstasy use [38, 39] and marijuana use [40, 41] with sexual risk-taking. However, studies in this area have not been consistent and further investigation into the association between ecstasy and risky sexual behavior including inconsistent condom use is warranted [42]. Similarly, the null findings for crack cocaine use and inconsistent condom use in the current study contrast with research in other settings among drug-using youth [43], suggesting more investigation is needed.

Study findings also show that female youth are significantly more likely to engage in unprotected sex, which is linked with complex interactions of gender inequality, power, and socio-structural context [44]. Our results indicate that condoms are inconsistently used among participants in stable relationships and who have multiple concurrent sex partners (38 % and 63 % at baseline, respectively). The positive relationship between stable relationships and inconsistent condom use aligns with previous research [45, 46], however, the association between multiple concurrent sex partners and inconsistent condom use is novel. These results point to the need for tailored gender-informed interventions to support consistent condom use among sexually active street-involved youth [47, 48].

There are a number of study limitations. The absence of a probability sample limits the ability of this study to generalize to other settings, although our extensive recruitment efforts resulted in a similar sample to those found in other studies of Vancouver street-involved youth [49, 50]. Self-report surveys are also vulnerable to recall and socially-desirable response biases [51]; however, under-reporting of illicit drug use and sexual practices are expected to bias our results to the null.