In this paper, we describe a cohort of injection drug users whose primary drug of choice was heroin. The demographic makeup of the cohort varied by age, with individuals <25 years old more likely to be female and white non-Hispanic. The cohort was comprised of individuals with fairly severe substance use, with the majority reporting daily injection use and daily heroin use. In addition, most reported experiencing withdrawal symptoms prior to their most recent injection equipment sharing event.
Injection drug use frequency did not vary by age; however, sharing of injection equipment did vary by age, with individuals <25 years old significantly more likely to share than individuals ≥25 years old. Our findings are consistent with those of most previous studies that examined the association between age and sharing. This association remained strong after adjusting for several risk factors of sharing including frequency of cocaine use, frequency of injection, and whether US born. We did not have information on age at initiation of injection drug use. However, it is reasonable to propose that younger individuals in our cohort on average initiated injection drug use more recently than did older individuals. Since people are often introduced to injection drug use by other users, one reason for the association between younger age and sharing may be that younger, newer users still rely on others’ help when injecting . Our finding that younger individuals who share were more likely than older individuals to report that they did not have their own drugs to inject is consistent with this reason. Use and abuse of prescription opioids is also high among young individuals and has emerged as an important introduction to heroin use and heroin injection . It is believed to be one of the reasons for the recent increase in injection drug use and HCV infection in young adults observed in Massachusetts and elsewhere .
While younger age was, overall, a strong risk factor for sharing, we found this association differed qualitatively depending on other factors that may be related to personal stressors or social instability. To our knowledge, this is the first study to have examined and identified potential effect modifiers of the association between age and sharing. Specifically, we found that the effect of younger age on sharing was substantial and significant among individuals who were born in the US, who did not have a history of receiving money for sex, or who did not have symptoms of PTSD. However, among those with one or more of these stressors, younger age was not associated with sharing. While we emphasize that these effect modification analyses and results are exploratory, they do suggest that, in the presence of other destabilizing factors in the lives of injection drug users, sharing is more prevalent among all ages, and the relative importance of younger age as a risk factor for sharing is diminished.
We found evidence that, in the presence of these stressors, the prevalence and frequency of sharing by older users was greater and more similar to that of younger users, than in the absence of these stressors. Previous research has found that older users or individuals with more extensive injection drug use experience are less likely to share than are younger users, either because they have learned how to inject on their own or because they have learned the health risks of sharing injection equipment . However, it could be that in the chronic presence of stressors such as PTSD or exchange sex, individuals are not able to develop the capacity to reduce their frequency of sharing as they get older. It may also be that individuals with these stressors have less stable or more frequently changing social networks, something which has also been associated with more risky drug using behavior [24, 25]. Factors specific to the social networks or life stressors of injection drug users may in part explain the conflicting results of previous studies regarding age and injection equipment sharing. For example in Stein et al. , where age was not significantly associated with sharing, all study participants met DSM-IV criteria for major depressive disorder, dysthymia, or persistent substance-induced mood disorder.
Our results revealed greater reporting of sharing by those born outside of the US. Most of the individuals in our cohort who were born outside of mainland US were born in Puerto Rico. Zerden et al.  recently showed that Puerto Ricans born and residing in Puerto Rico were more likely to share needles than were Puerto Ricans born in the US and residing in Massachusetts.
Previous research has found that young injection drug users have lower rates of HCV testing than older users . We did not observe the same finding in our cohort. This may be due to the fact that these individuals were identified in a clinical setting with minimal financial barriers to care, and may, therefore, have had more access to health care and testing services than other populations of injecting drug users.
It is worth commenting that syringe and needle exchange and legal access to over-the-counter purchases have been successful public health strategies in Massachusetts; however, in our study, the majority of individuals who reported sharing said the equipment used in their most recent sharing event was obtained at either needle exchange or pharmacy, and most individuals who were tested for HCV reported testing positive. Access to clean equipment for these individuals has, therefore, not translated into sufficient protection against HCV infection. Strategies to address the precursors of sharing are critically needed.
Our study has several limitations related to the cross-sectional nature of the analyses. The lower prevalence of sharing observed for older users might not be due to safer injecting practices, but may instead be due to an overall reduction in substance use or in injection frequency over time, thereby reducing the opportunity for sharing. However, in our cohort, the frequency of injection drug use as well as the frequency of heroin use (the drug most individuals reported injecting) did not differ for older and younger users. Alternatively, the lower prevalence of sharing could be due to a survivor effect, where older individuals with years of sharing injection drug equipment do not survive and are not represented in our cohort. We are not able to assess this likelihood in the present study.
We did not have information on age at initiation of injection drug use and, therefore, cannot confirm that younger users initiated injection drug use more recently than older users. We did not collect information on the social networks of our participants; the potential influence of social networks on the association between age and sharing should be specifically assessed in future investigations. Our examination of effect modification was exploratory, as the study was not designed to examine effect modification and was, therefore, not optimally powered to identify statistically significant modifiers. Future studies with larger study samples should examine the same and additional effect modifiers related to social stress and instability, such as stressful life events and social support.
Also, since our cohort was comprised of injection drug users registered for health care at a busy urban emergency department, they are not necessarily representative of all injection drug users in Massachusetts. We did not conduct HCV testing and instead relied on self-reported information on HCV testing practices, which may mean that the observed prevalence of HCV testing was subject to recall bias. However, we did conduct HIV testing, and HIV test results were highly consistent with individuals’ self-reports of their HIV serostatus, suggesting that biased reporting of HCV testing patterns was minimal (data not shown).