How do drug market changes affect characteristics of injecting initiation and subsequent patterns of drug use? Findings from a cohort of regular heroin and methamphetamine injectors in Melbourne, Australia

• Few studies have examined the impacts of changes in drug market characteristics on injecting initiation experiences.
• Participants who initiated injecting during a setting of low heroin availability were twice as likely to initiate injecting using a drug other than heroin.
• In addition to influences on first drug injected, the drug market period in which injecting initiation occurred also influenced some aspects of subsequent drug use.

BackgroundChanges in drug market characteristics have been shown to affect drug use patterns but few studies have examined their impacts on injecting initiation experiences and subsequent patterns of injecting drug use (IDU).MethodsWe collected data on self-reported injecting initiation experiences and past-month patterns of IDU from 688 regular heroin and methamphetamine injectors in Melbourne, Australia, who initiated injecting across three different drug market periods (prior to the Australian heroin shortage (‘high heroin’)/immediately following the shortage (‘low heroin’)/‘contemporary’ markets (fluctuating heroin and methamphetamine availability)). We used univariable and multivariable logistic regression to examine the relationship between period of injecting initiation and first drug injected, and multinomial logistic regression for the relationship between period of injecting initiation and current injecting patterns.Results425 participants (62%) reported initiating injecting in the high heroin period, 146 (21%) in the low heroin period, and 117 (17%) in the contemporary period. Participants who initiated injecting during the low heroin period were twice as likely to initiate injecting using a drug other than heroin (AOR: 1.94, 95% CI: 1.27–2.95). The most common patterns of drug use among study participants in the month preceding interview were polydrug use (44%) and primary heroin use (41%). Injecting initiation period was either non-significantly or weakly associated with current drug use pattern, which was more strongly associated with other socio-demographic and drug use characteristics, particularly self-reported drug of choice.ConclusionThe drug market period in which injecting initiation occurred influenced the first drug injected and influenced some aspects of subsequent drug use. In the context of highly dynamic drug markets in which polydrug use is common there is a need for broad harm reduction and drug treatment services which are flexible and responsive to changing patterns of drug use.