For the past two decades, IDUs have accounted for an important proportion of new human immunodeficiency virus (HIV) infections in most developed countries. High-risk injection practices, particularly needle borrowing [1–7], and risky sexual practices [2,5,8–10] have contributed to HIV transmission among IDUs. To slow the epidemic, most countries have deployed harm reduction efforts such as community outreach, needle exchange programmes (NEPs), supervised injection sites, and substitution treatment; however, worldwide implementation is not yet optimal. These HIV prevention programmes have been found to be successful in decreasing HIV transmission [11–14]. Concurrently, HIV seroconversion among IDUs has declined or stabilized in North America [5,15–22] and Western Europe [23,24]. However, HIV incidence rates remain high in certain regions. In eastern central Canada, HIV incidence among IDUs from the SurvUDI network – an HIV and related high-risk behaviour surveillance programme for IDUs – was 3.6 per 100 person-years in 1998; it decreased to 1.8 per 100 person-years in 2002 and then rose to 2.7 per 100 person-years in 2006 .
Since the beginning of the epidemic, determinants of HIV acquisition may have changed as a result of sustained prevention programmes. However, few studies have looked at changes in risk factors for HIV incidence over time. A prospective cohort study of HIV-negative IDUs in Baltimore indicated that drug use practices at high risk for HIV transmission decreased between 1988 and 1998, and then increased slightly between 2003 and 2004 [5,18]. Female participants had a lower risk of seroconversion than male participants in the 1993–1998 period and a higher risk in the 1999–2004 period, although the interaction between sex and time was not statistically significant [5,18]. Despite the importance of understanding the evolution of risk factors driving the HIV epidemic, few studies have documented temporal changes in risk factors for HIV incidence among IDUs.
The present study investigates temporal trends (1995–2009) in HIV incidence rates among IDUs from the SurvUDI network. It also aims to assess risk factors associated with HIV seroconversion and to verify if these risk factors changed over time.
The complete methodology of the SurvUDI study has been described elsewhere . Briefly, the SurvUDI study is based on a second-generation surveillance programme for HIV and risk behaviours among IDUs in eastern central Canada. The ongoing surveillance network was implemented in 1995 and targeted hard-to-reach, mostly out-of-treatment IDUs. Eligibility criteria include being aged 14 and older, injecting at least once within the past 6 months, speaking French or English and being able to provide informed consent. Participants were recruited in urban areas, including Montréal and neighbouring South Shore, Québec City, the Hull-Ottawa region, and five semi-urban areas of the province of Québec. Overall, 88.4% of participants were recruited in NEPs. Others were recruited in drop-in centres, detention centres, detoxification clinics, and rehabilitation programmes. Participation includes an interviewer-administered questionnaire and collection of a saliva sample using the Orasure device (Bethlehem, Pennsylvania, US) for HIV antibody testing. Participants are encouraged to complete interviews at 6-month intervals. They are given a stipend ranging from CAN$5.00 to $10.00 at the end of each study visit. All procedures have been approved by the ethics committee of the Centre hospitalier affilié universitaire de Québec.
On the basis of literature, study variables included cocaine as the drug the most often injected [2,5,25,26], needle borrowing [1,2,4,6,7,10], sex work (i.e. sex in exchange for money or other goods or services) [2,5,8,9], and injection frequency [2–5,10,25,26]. Questions about the first three behaviours referred to the 6 months prior to the interview, whereas injection frequency (number of injections per day) was documented for the previous month to reduce recall bias, thus ensuring optimum precision. Sociodemographic characteristics included sex, age and the region where the interview took place (urban or semi-urban).
Collected oral fluid samples were kept at 4°C and shipped within 2 weeks to the Laboratoire de Santé Publique du Québec (LSPQ; Institut national de Santé Publique du Québec), where they were centrifuged upon reception. The extracted liquid was kept at −20°C for a maximum of 6 weeks until analysis. The presence of HIV antibodies was assessed by enzyme immunoassay (EIA) using HIV-1 Vironostika Microelisa System (bioMérieux, Durham, North Carolina, USA) from 1995 to 2007 and Detect HIV-1/2 (Adaltis, Montréal, Québec) thereafter. Samples were considered negative for HIV if EIA results were less than 75% of the cut-off value. Sample results that were greater than 75% of the cut-off value were retested in duplicate. A sample was deemed positive if at least two out of three results were greater than the cut-off value.
To verify similarities between HIV-negative repeaters and HIV-negative participants who completed a single visit, baseline characteristics and HIV risk behaviours were compared using Pearson's chi-squared tests. Frequencies of risk factors reported among HIV-negative repeaters for each period were compared using the score test with the generalized estimating equations method . This analysis considers the dependency of observations. In total, 911 (58%) participants who contributed to the person-years in 2003–2009 also contributed in 1995–2002. All participants could participate in both time periods; however, participants were excluded in the 2003–2009 period if they had seroconverted before 2003.
Only HIV-negative repeaters at baseline were considered when assessing HIV incidence rates. To take variable periods of observation into account, we computed time trends in HIV incidence with a method that assigns a rising probability of seroconversion to each day between a participant's last negative and first positive HIV test result , rather than assuming that seroconversion had occurred at the midpoint between the last negative HIV test and the first positive HIV test result. Incidence by time period (1995–2002 and 2003–2009) was computed with proportional attribution of cases over time, and the difference was assessed. Given that many participants contributed to both time periods, differences between incidence rates over the two periods were statistically compared using a bootstrap approach .
Bivariate and multivariate Cox proportional hazard regression analyses were carried out to assess predictors of HIV acquisition. Behavioural variables and age were treated as time-dependent variables. Independent variables whose distribution changed significantly between the time periods or for which HIV incidence rates changed significantly over time were considered in the analyses. Interactions between each covariate and time period were assessed and kept in the model if they were statistically significant at a P value 0.15 or less. All analyses were carried out using SAS statistical suite software version 9.2 (SAS Institute Inc., Cary, North Carolina, USA).
Characteristics of injection drug users
A total of 11 731 IDUs were recruited in the SurvUDI study between January 1995 and June 2009. Men comprised 75.5% of the sample, and median age at recruitment was 33 years for men and 28 years for women. At baseline, the proportion of participants reporting cocaine as the most often injected drug in the past 6 months was 71.1%; it was 24.5% for opioids and 4.4% for other drugs [including phencyclidine, crack, steroids, and others]. Furthermore, 34.5 and 16.6% of IDUs reported borrowing needles and engaging in sex work in the past 6 months, respectively, and 27.7% had injected at least once a day in the past month.
At baseline, HIV prevalence was 11.9%, leaving 10 231 HIV-negative participants potentially eligible for an incidence study. Among them 2903 (28.4%) had participated in SurvUDI more than once, with a mean follow-up time of 3.8 years (SD = 3.3) and 3.7 visits (SD = 2.5). HIV-negative IDUs who had participated only once were slightly but significantly different from repeaters (Table 1). They were more likely to be older and male, and, except for cocaine use, were less likely to report HIV risky behaviours.
On the basis of information collected at the first visit of each time period, Table 2 shows that the proportion of HIV-negative repeaters reporting frequent injection (once a day or more) was higher in the second period. Other risky behaviours were more prevalent in the first period. HIV-negative repeaters were significantly older in the second period. Further analyses show that some of the temporal differences in high-risk behaviours vary by age groups: cocaine use and sex work significantly declined only among older participants (84.3 vs. 70.2%, P < 0.0001 and 18.0 vs. 13.6%, P = 0.0004), whereas frequent injection increased only among younger participants (26.5 vs. 39.7%, P < 0.0001).
Temporal trends in HIV incidence
Of the 2903 HIV-negative repeaters, 276 seroconverted during follow-up. These IDUs accumulated 10 106.6 person-years of infection-free follow-up, for a global incidence rate of 2.7 per 100 person-years [95% confidence interval (CI) 2.4–3.1]. More than half (58%) of participants who contributed to the person-years in 2003–2009 also contributed in 1995–2002. Overall, HIV incidence declined significantly between time periods (Table 3). HIV incidence decreased significantly among men (P = 0.0003), whereas a significant (P = 0.0243) increase in HIV incidence was observed among younger IDUs (<25 years), with the rate doubling in the second period. In contrast, incidence declined by 44% among older IDUs.
Predictors of HIV seroconversion by time period
Interaction analyses showed that the association between HIV incidence and some risk factors varied by time period at a P value less than 0.15. These risk factors are injecting at least once a day, age, sex work, and being recruited in an urban area. A multivariate Cox proportional hazard model was computed with all covariates and their interaction terms for both time periods. HIV acquisition was more likely to occur among men, needle borrowers and IDUs injecting cocaine most often, for both time periods (Table 4, see subsection ‘No significant interaction with time’). Being older, injecting every day and being recruited in an urban centre were significant risk factors for HIV incidence between 1995 and 2002 only. Finally, sex work emerged as a significant predictor of HIV seroconversion for the 2003–2009 period.
On the basis of surveillance data, the study shows that HIV incidence among IDUs decreased in eastern central Canada from 3.1 per 100 person-years in 1995–2002 to 2.2 per 100 person-years in 2003–2009. These findings are consistent with a recent prospective cohort study conducted among IDUs in Montréal which indicates a decreasing HIV incidence between 1992 (3.5 per 100 person-years) and 2008 (1.8 per 100 person-years) . Findings are also coherent with the decline observed from the mid-1990s to the mid-2000 among IDUs elsewhere in North America [5,15,16,18].
Analyses of risk behaviours associated with HIV incidence by time period showed that needle borrowing and cocaine injection remained significant predictors for HIV infection over time. It is noteworthy that, despite a significant decrease between the first and second time periods, almost a quarter of SurvIDU repeater participants still reported borrowing needles. Despite a slight decrease in this period, cocaine remained the drug most often injected. This is worrisome given that cocaine injectors tend to have multiple binge periods that typically involve erratic behaviours, thus enhancing risks of unsafe practices and blood exposures .
Whereas daily injection increased the risk of HIV infection by 50% in the 1995–2002 period, it no longer predicted HIV incidence in the 2003–2009 period. This may be explained by the increasing number of NEPs, pharmacies and centres distributing clean needles and syringes since 1996 in Québec , which improved access to sterile injection equipment. Consequently, despite the increasing number of regular IDUs (injecting ≥1 per day), the overall HIV risk related to frequent injection may have declined. However, a study assessing clean needle/syringe needs in Montréal (the largest city in the surveillance network) indicates that the amount of clean injection material distributed is still insufficient . According to a recent review on worldwide needle-syringe distribution coverage, it is estimated that 46 needle/syringes are distributed to each Canadian IDU per year. Such coverage is low in comparison to several European countries and Australia, where 213 needle/syringes per IDU are distributed yearly . Increased efforts to improve coverage are needed.
Our results show a significant interaction between age and time period, with older age no longer predicting HIV incidence in the 2003–2009 period. A decline in HIV incidence among older IDUs and a rise among younger ones were also observed. A possible explanation for the overall decrease and levelling between age groups is that the ‘repeaters’ cohort is getting older. In fact, IDUs aged under 25 represented 29% of person-years in 1995–2002 compared to 14% in 2003–2009 (Table 3), and more than half of participants contributed to person-years in both time periods. It is plausible that most high-risk IDUs were infected soon after recruitment in the cohort, leading to a decrease in incidence over time because of a saturation effect . Moreover, analyses showed that some temporal differences in high-risk behaviours varied by age groups. Cocaine use and sex work declined among older participants during the study period and frequent injection increased among younger ones. These changes could be at stake in the increasing HIV incidence rate observed among young IDUs. One explanation might reside in treatment optimism, more often reported in men who have sex with men (MSM) but also observed among IDUs . We must increase our knowledge of the consumption patterns of young drug users who initiate injection in a context of an evolving drug market that impacts drug availability and routes of drug administration which, in turn, may influence drug users’ HIV risk behaviours [36–38]. Consequently, there is a concern that young IDUs will relax their safer practices.
Despite the decrease in HIV incidence among men and its stability among women over time, multivariate models did not show a significant interaction between sex and time period, but rather an enhanced HIV risk among men throughout the study. The study findings are consistent with those of an open cohort of IDUs carried out in Montréal where men were twice as likely than women to seroconvert during follow-up . However, other studies have shown inconsistent results regarding sex and HIV risk among IDUs [5,17,18,23,25]. Whereas sex work was not a significant determinant of HIV incidence in the 1995–2002 period, it was found to be an independent risk factor in the 2003–2009 period. Although this association could be due to residual confounding of injection practices, this result may reflect a real sexual risk. A context of sexual ‘prevention fatigue’  may have contributed to a rise in risky sexual practices among IDUs. It is possible that high-risk sexual practices are likely to lead to HIV transmission among IDU sex workers. However, there are no recent data on sexual transmission of HIV in Québec among IDUs involved in sex work. Therefore, it is premature to reach a conclusion regarding the causes underlying this association. It is important to broaden our understanding of the contexts and practices associated with seroconversion among IDU sex workers. Finally, between 1995 and 2002, IDUs recruited in urban areas were more at risk of seroconverting compared to IDUs recruited in semi-urban areas; however, this association disappeared in the period 2003–2009. The reasons for these findings are unknown and warrant further study.
This study presents some limitations. The observed HIV incidence rate might be overestimated compared with the overall IDU population. HIV-negative repeaters were more likely to report HIV risk behaviours at baseline in comparison to IDUs who completed a single visit. Therefore, findings should only be generalized to IDUs with more problematic behaviours who attend harm reduction community-based resources. Furthermore, self-reported behavioural measures may lead to social desirability and recall biases that might have caused an underestimation of the proportion of participants engaging in risky behaviours. However, previous studies have shown the validity and reliability of IDUs’ self-reported behaviours [39,40].
In this study on IDUs from eastern central Canada mainly recruited in harm reduction programmes, we observed a decline in HIV incidence in the period 1995–2009. Although needle borrowing and cocaine injection remain the most important risk factors explaining HIV acquisition in this population, the contribution of sex work became significant over time and HIV incidence increased among young IDUs in recent years. Interventions targeting IDUs must be maintained and diversified, including efforts to prevent risky sexual practices, while ensuring optimal coverage by harm reduction programmes. Moreover, new IDUs, in particular the youngest ones, should constitute a high priority group for public health research and intervention.
All authors contributed to the study. C.B. and M.A. undertook the analysis and É.R. and I.R. drafted the article. É.R., C.M., P.L., C.C., R.P. and M.A contributed to the design of the study. All authors provided feedback on drafts and approved the final version.
The SurvUDI network was supported by the Public Health Agency of Canada and the Ministère de la Santé et des Services Sociaux du Québec (MSSS).
The authors would like to acknowledge the contribution of the SurvUDI group, the staff from all centres where recruitment is conducted and injection drug users for giving their time.
Work was carried out at Faculté de Médecine et des Sciences de la Santé, Université de Sherbrooke, Service de Toxicomanie, Longueuil, Québec, Canada.
Conflicts of interest
There are no conflicts of interest to disclose.
1. Chaisson RE, Moss AR, Onishi R, Osmond D, Carlson JR. Human immunodeficiency virus-infection in heterosexual intravenous-drug-users in San Francisco
. Am J Public Health
2. Hankins C, Alary M, Parent R, Blanchette C, Claessens C. the SurvUDI Working GroupContinuing HIV transmission among injection drug users in Eastern central Canada: the SurvUDI study, 1995 to 2000
. J Acquir Immune Defic Syndr
3. Marmor M, Des Jarlais DC, Cohen H, Friedman SR, Beatrice ST, Dubin N, et al. Risk factors for infection with human immunodeficiency virus among intravenous drug abusers in New York City
4. Miller CL, Kerr T, Frankish JC, Spittal PM, Li K, Schechter MT, et al. Binge drug use independently predicts HIV seroconversion among injection drug users: implications for public health strategies
. Subst Use Misuse
5. Nelson KE, Galai N, Safaeian M, Strathdee SA, Celentano DD, Vlahov D. Temporal trends in the incidence of human immunodeficiency virus infection and risk behavior among injection drug users in Baltimore, Maryland, 1988–1998
. Am J Epidemiol
6. Schoenbaum EE, Hartel D, Selwyn PA, Klein RS, Davenny K, Rogers M, et al. Risk factors for human immunodeficiency virus infection in intravenous drug users
. N Engl J Med
7. Vlahov D, Munoz A, Anthony JC, Cohn S, Celentano DD, Nelson KE. Association of drug injection patterns with antibody to human immunodeficiency virus Type-1 among intravenous drug users in Baltimore, Maryland
. Am J Epidemiol
8. Celentano DD, Latimore AD, Mehta SH. Variations in sexual risks in drug users: emerging themes in a behavioural context
. Curr HIV/AIDS Rep
9. Kral AH, Bluthenthal RN, Lorvick J, Gee L, Bacchetti P, Edlin BR. Sexual transmission of HIV-1 among injection drug users in San Francisco, USA: risk factor analysis
10. Strathdee SA, Galai N, Safaiean M, Celentano DD, Vlahov D, Johnson L, et al. Sex differences in risk factors for HIV seroconversion among injecting drug users
. Arch Intern Med
11. Gowing L, Farrell M, Bornemann R, Sullivan LE, Ali R. Substitution treatment of injection opioid users for prevention of HIV infection (Review)
. Cochrane Database Syst Rev
12. Palmateer N, Kimber J, Hickman M, Hutchinson S, Rhodes T, Goldberg D. Evidence for the effectiveness of sterile injecting equipment provision in preventing hepatitis C and human immunodeficiency virus transmission among injecting drug users: a review of reviews
13. Van Den Berg C, Smit C, Van Brussel G, Coutinho R, Prins M. Full participation in harm reduction programmes is associated with decreased risk for human immunodeficiency virus and hepatitis C virus: evidence from the Amsterdam Cohort Studies among drug users
14. Vlahov D, Robertson AM, Strathdee SA. Prevention of HIV infection among injection drug users in resource-limited settings
. Clin Infect Dis
15. Des Jarlais DC, Marmor M, Friedmann P, Titus S, Aviles E, Deren S, et al. HIV incidence among injection drug users in New York City, 1992–1997: evidence for a declining epidemic
. Am J Public Health
16. Hall HI, Song R, Rhodes P, Prejean J, An Q, Lee LM, et al. Estimation of HIV incidence in the United States
17. Kral AH, Lorvick J, Gee L, Bacchetti P, Rawal B, Busch M, et al. Trends in human immunodeficiency virus seroincidence among street-recruited injection drug users in San Francisco, 1987–1998
. Am J Epidemiol
18. Mehta SH, Galai N, Astemborski J, Celentano DD, Strathdee SA, Vlahov D, et al. HIV incidence among injection drug users in Baltimore, Maryland (1988–2004)
. J Acquir Immune Defic Syndr
19. Parent R, Alary M, Morissette C, Roy É, Leclerc P, Allard P-R. Surveillance
des maladies infectieuses chez les utilisateurs de drogue par injection. Épidémiologie du VIH de 1995 à 2008. Épidémiologie du VHC de 2003 à 2008. Institut national de santé publique du Québec. 2009. (Accessed September 23, 2010, at http://www.inspq.qc.ca/pdf/publications/1021_SurvMalInfecUDI_VIHVHC2008.pdf
21. Santibanez SS, Garfein RS, Swartzendruber A, Purcell DW, Paxton LA, Greenberg AE. Update and overview of practical epidemiologic aspects of HIV/AIDS among injection drug users in the United States
. J Urban Health
22. Bruneau J, Daniel M, Abrahamowicz M, Zang G, Lamothe F, Vincelette J. Trends in human immunodeficiency virus incidence and risk behavior among injection drug users in Montreal, Canada: a 16-year longitudinal study
. Am J Epidemiol
23. Hurtado Navarro I, Alastrue I, Del Amo J, Santos C, Ferreros I, Tasa T, et al. Differences between women and men in serial HIV prevalence and incidence trends
. Eur J Epidemiol
24. Sabbatini A, Carulli B, Villa M, Corrêa Leite ML, Nicolosi A. Northern Italian Seronegative Drug Addicts (NISDA) study. Recent trends in the HIV epidemic among injecting drug users in Northern Italy, 1993–1999
25. Spittal PM, Craib KJ, Wood E, Laliberté N, Li K, Tyndall MW, et al.Risk factors for elevated HIV incidence rates among female injection drug users in Vancouver
26. Tyndall MW, Currie S, Spittal P, Li K, Wood E, O'Shaughnessy MV, et al. Intensive injection cocaine use as the primary risk factor in the Vancouver HIV-1 epidemic
27. Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models
28. Kitayaporn D, Uneklabh C, Weniger BG, Lohsomboon P, Kaewkungwal J, Morgan WM, et al. HIV-1 incidence determined retrospectively among drug users in Bangkok, Thailand
29. Efron B, Tibshirani RJ. An introduction to the bootstrap.
Monographs on Statistics and applied probability 57. Florida: Chapman & Hall/CRC; 1993.
30. Bourgois P, Bruneau J. Needle exchange, HIV infection, and the politics of science: Confronting Canada's cocaine injection epidemic with participant observation
. Med Anthropol
31. Noël L, Cloutier R. Statistiques sur les services relatives aux programmes de prévention du virus de l’immunodéficience humaine (VIH) et des hépatites B et C auprès des personnes utilisatrices de drogues par injection au Québec. Institut national de santé publique, Québec. Bibliothèque et Archives nationales du Québec. 2009. ISBN: 1921–9377. (Accessed 26 April, 2011, at http://www.inspq.qc.ca/pdf/publications/1022_StatServicesProgVIHVHBVHC_07-08.pdf
32. Leclerc P, Morissette C, Tremblay C. Le matériel stérile d’injection: combien faut-il en distribuer pour répondre aux besoins des UDI de Montréal? Direction de la santé publique. Agence de la santé et des services sociaux de Montréal. 2006. ISBN: 2-89494-499-3 (Accessed September 30, 2011, at http://www.santepub-mtl.qc.ca/Publication/pdfmi/materielinjection.pdf
33. Mathers BM, Degenhardt L, Ali H, Wiessing L, Hickman M, Mattick RP, et al. HIV prevention, treatment, and care services for people who inject drugs: a systematic review of global, regional, and national coverage
34. Winkelstein W Jr, Samuel M, Padian NS, Wiley JA, Lang W, Anderson RE, et al. The San Francisco Men's health study: III. Reduction in human immunodeficiency virus transmission among homosexual/bisexual men 1982–86
. Am J Public Health
35. Tun W, Celentano DD, Vlahov D, Strathdee SA. Attitudes towards HIV treatments influence unsafe sexual and injection practices among injection drug users
36. Ciccarone D. Heroin in brown, black and white: structural factors and medical consequences in the US heroin market
. Int J Drug Policy
37. Ciccarone D, Bourgois P. Explaining the geographical variation of HIV among injection drug users in the United States
. Subst Use Misuse
38. Sterk CE, Elifson KW. Fluctuating drug markets and HIV risk taking: female drug users and their relationship with drug markets
. Med Anthropol
39. Darke S. Self-report among injecting drug users: a review
. Drug Alcohol Depend
40. Goldstein MF, Friedman SR, Neaigus A, Jose B, Ildefonso G, Curtis R. Self-reports of HIV risk behavior by injecting drug users: are they reliable?