Street involvement among youth has been recognized as an important and growing public health problem worldwide . This reality is difficult to define and may present differently depending on the social environment where it takes place. This explains the great variety of labels that are used to designate these youth. In the developing world, street youth are usually categorized as ‘home-based’ or ‘street-based’ depending on their level of street involvement and the degree of family affiliation. In the developed world, terms like street youth, homeless youth, runaways or throwaways are used, underlining the fact that these youth are often estranged from their homes and their families. Despite this diversity of labels, these youth all face precarious living conditions, including poverty and residential instability, combined with varying degrees of involvement in the street economy. This often translates into a lifestyle based on survival, and the various risk behaviours it engenders, such as injection drug use [2–9] and prostitution [2–6,9,10], put street youth at increased risk for numerous health problems. HIV infection is one of the many health problems that street youth have to face.
To date, the HIV situation among street youth has been described in terms of prevalence but never of incidence. Two South American studies reported HIV prevalence of 3.2 and 4.6% [11,12]. In the United States, estimates of HIV prevalence varied widely according to recruitment site, city and study period, from 0 to 11.5% [4,13–15]. In Canada, prevalence of 1.9 and 2.2% were observed in two major cities [5,6]. Factors most frequently identified as associated with prevalence were injecting drug use [5,6,12], homosexual activities among males [5,13,14], and prostitution [4–6,12]. Other identified correlates of infection were having had another sexually transmitted disease , crack use , unprotected vaginal sex , sex with an injection drug user , older age [5,6], and birth outside Canada .
The prospective cohort study described here was conducted to determine HIV incidence among street youth in Montreal and to identify risk factors for HIV infection.
The study was initiated in Montreal, Canada in 1995. The complete methodology has been described previously . Briefly, criteria for entry in the study were being ‘street-active', 14–25 years of age, English or French speaking, and being able to provide informed consent and to complete a questionnaire. Youth were considered ‘street-active’ if they had, in the last year, either regularly used the services of street youth agencies or been without a place to sleep more than once. Study interviewers recruited participants through regular visits to all major street youth agencies in Montreal. These agencies offered mainly drop-in centre services, shelter and outreach services on the street.
Participants were interviewed twice a year; they completed a 45-minute interviewer-administered questionnaire covering sociodemographic characteristics, alcohol and drug use, and sexual behaviours, and provided two samples of gingival exudate for HIV antibody testing. Each visit was financially compensated (CAD $20). Ethical approval was provided by the Institutional Review Board of the Faculty of Medicine, McGill University.
Saliva samples were kept at 4°C and sent weekly to the provincial laboratory, where they were extracted by centrifugation and frozen at −20°C. Specimens were tested for the presence of HIV antibodies by Vironostika HIV-1 enzyme immunoassay (EIA) (Organon-Teknika Inc., Scarborough, Ontario, Canada). Reactive specimens were tested in duplicate; repeatedly reactive specimens were considered positive.
The HIV incidence analysis was restricted to cohort participants who were HIV-negative at study entry and had completed at least one follow-up interview by 30 September 2000. Incidence was calculated as the number of participants who first tested HIV-positive during follow-up divided by the total person-time under observation. Person-time was defined as the interval between enrolment and either the most recent follow-up visit (for non-seroconverters) or the date of seroconversion (for seroconverters). The date of seroconversion was defined as the midpoint between the two visits with the last HIV-negative and the first HIV-positive test results. A 95% confidence interval (CI) for the global incidence estimate was calculated using the Poisson distribution. Risk ratios and 95% CI values of potential predictors of seroconversion were determined using Cox proportional hazard regression. In univariate analyses, crude incidence rates of infection were calculated as the number of youth who were exposed (or non-exposed) at the time of infection divided by the person-time attributed to the exposed (or non-exposed) category. All variables with P values ≤ 0.20 in univariate analyses were included in the multivariate Cox model. Variables with P values ≤ 0.05 were considered statistically significant.
Youth characteristics considered as potential predictors were sex, age, injection drug use, being a boy having male sexual partners (including clients), and survival sex (defined as receiving money, gifts, drugs, a place to sleep or something else in exchange for sexual activities). All independent variables except the sex were treated as time-dependent, covering the preceding 6 months and re-assessed at each interview.
From 24 January 1995 to 30 September 2000, 1013 youth were recruited into the cohort, with a refusal rate estimated at 12%. Fourteen youth (2 girls, 12 boys) tested HIV positive at enrolment, for an HIV prevalence of 1.4% (95% CI, 0.8–2.4%). The HIV prevalence at entry was stable across the 6 recruitment years, varying between a minimum of 0.9% (95% CI, 0.1–4.7%) in 1998 and a maximum of 1.7% (95% CI, 0.4–4.9%) in 1999.
At baseline, the mean age of the 863 participants selected for the HIV incidence analysis was 19.8 years (SD 2.5). Two thirds (66.7%) were boys; the majority (95.1%) were born in Canada, and almost all (96.1%) had ever been without a place to sleep. Substance use was prevalent, with 97.0% having ever used marijuana/hashish, 91.3% hallucinogens (including LSD, PCP and mushrooms), 81.5% cocaine/crack/freebase, and 41.3% heroin. Near half (47.2%) had ever injected drugs. Almost all participants (99.0%) had engaged in some type of sexual activity; 22.1% had ever had same-sex partners (28.9% girls and 18.8% boys); 25.7% had ever engaged in survival sex, and 37.8% had a history of sexual abuse.
As of 30 September 2000, the 863 participants had cumulated 2327 person-years of follow-up and 16 (5 girls, 11 boys) had seroconverted, for an incidence rate of HIV infection of 0.69 per 100 person-years (95% CI, 0.39–1.11). Fifteen seroconverters were born in Canada and one in Western Europe; none had ever received blood/blood products outside Canada.
Table 1 presents the results of the Cox regression analyses. In univariate analyses, injection drug use and involvement in survival sex were significantly associated with HIV incidence. When both variables were included within the same model, injection drug use was the only variable that remained statistically significant.
This is the first study to measure HIV incidence in street youth. The observed rate of 0.69 per 100 person-years is in the same range as those reported for cohorts of gay and bisexual men in Montreal (0.56 per 100 person-years)  and non-injection drug using gay and bisexual men in Vancouver (1.0 per 100 person-years) .
Injection drug use was the only identified predictor of HIV seroconversion among street youth. The observed incidence rate in cohort participants reporting drug injection was 1.72 per 100 person-years (95% CI, 0.89–2.99). This is lower than the rates of 2.6% per year observed in an American cohort of injection drug users aged 18 to 29 years  and 4.4 per 100 person-years observed in a subset of participants aged 24 years or less in a Canadian cohort of injection drug users . This difference may result from several factors, including the duration and penetration of the HIV epidemic in the local population and the different patterns of social mixing among young injecting drug users . We could not examine the association between specific needle-sharing behaviours and seroconversion because of the small number of cases. However, at study entry, 53% of youth reported having ever injected with a syringe already used by someone else and, from 1995 to 2000, this proportion remained constant. This result suggests that the potential for an HIV outbreak continues to exist among young injecting drug users in Montreal, particularly in the context of the constantly high HIV incidence (6.0 per 100 person-years) recently reported for the ‘general’ population of injecting drug users in Montreal .
It is of particular interest that the annual cross-sectional analysis has shown stable HIV prevalence over the study period despite an incidence rate greater than zero. This highlights the importance of a longitudinal study design that measures incidence to truly capture the dynamics of an epidemic . It also suggests that perhaps the period during which youth remain on the streets is relatively short.
Limitations of this study must be taken into consideration in the interpretation of the results. First, the small number of seroconversions may have resulted in a limited power to detect predictors other than injection. It also precluded the possibility of separate analyses for males and females, which, on the basis of recent literature, could have been relevant [24,25]. A second area of concern may be the generalizability of the results. However, cohort participants should be representative of the larger Montreal street youth population given that recruitment was conducted in all major street youth organizations and that the refusal rate at recruitment was low. Moreover, as a recent survey has shown, most of the homeless population of Montreal (over 90%) attend community organizations offering services to homeless . Third, this study relies on self-reported information, which may be influenced by social desirability and recall biases. To reduce the social desirability bias, interviews were conducted in the study office, interviewers were unaffiliated with service agencies, and repeated assurances of confidentiality were given. The recall bias should have been reduced by the prospective data collection and the interval of 6 months between questionnaires. Finally, another concern could be selective losses to follow-up, which may have biased observed results. However, a comparison of the HIV-negative subjects lost to follow-up after their initial interview and those retained in the analysis showed that both groups were comparable at study entry with regard to the five variables included in the regression analyses. Also, the low attrition rate (5.7 per 100 persons-years) that was achieved through intensive follow-up procedures should have minimized this bias even further.
To decrease HIV incidence among street youth, prevention of initiation into drug injection must become a public health priority. Possible strategies may include more accessible drug treatment services for youth engaged in intensive drug use and targeted preventive messages to improve knowledge about different routes of drug administration and their consequences. For youth already injecting, interventions favouring alternatives to injection may be an option. However, the innocuity of the various routes of drug administration on a person's social and physical health has to be further documented.
The authors would like to thank the street youth and street youth agencies that collaborated in this study and all members of the research team.
Sponsorship: This study was supported by a grant from the Medical Research Council of Canada, Health Canada, the Fonds de recherche en santé du Québec, and the Ministère de la santé et des services sociaux du Québec.
1.WHO/OMS/PSA. A Two-way Street? Report on Phase II of the PSA Street Children Project.
Geneva: World Health Organization Division of Mental Health and Prevention of Substance Abuse; 1996.
2.Clatts MC, Davis WR,Sotheran JL, Atillasoy A. Correlates and distribution of HIV risk behaviors among homeless youths in New York City: implications for prevention and policy. Child Welfare
3.Gleghorn AA, Marx R, Vittinghoff E, Katz MH. Association between drug use patterns and HIV risks among homeless, runaway, and street youth in Northern California. Drug Alcohol Depend
4.Pfeifer RW, Oliver J. A study of HIV seroprevalence in a group of homeless youth in Hollywood, California. J Adolesc Health
5.DeMatteo D, Major C, Block B, Coates R, Fearon M, Goldberg E, et al
. Toronto street youth and HIV/AIDS: prevalence, demographics, and risks. J Adolesc Health
6.Roy É, Haley N, Leclerc P, Lemire N, Boivin J-F, Frappier J-Y, et al
. Prevalence of HIV infection and risk behaviours among Montreal street youth. Int J STD AIDS
7.Smart RG, Adlaf EM. Substance use and problems among Toronto street youth. Br J Addict
8.Ochnio JJ, Patrick D, Ho M, Talling DN, Dobson SR. Past infection with hepatitis A virus among Vancouver street youth, injection drug users and men who have sex with men: implications for vaccination programs. Can Med Assoc J
9.Kipke MD, O'Connor S, Palmer R, MacKenzie RG. Street youth in Los Angeles. Profile of a group at high risk for human immunodeficiency virus infection. Arch Pediatr Adolesc Med
10.Rotheram-Borus MJ, Koopman C, Ehrhardt AA. Homeless youths and HIV infection. Am Psychol
11.Avila MM, Casanueva E, Piccardo C, Liberatore D, Cammarieri G, Cervellini M, et al
. HIV-1 and hepatitis B virus infections in adolescents lodged in security institutes of Buenos Aires. Pediatr AIDS HIV Infect
12.Zanetta DMT, Strazza L, Azevedo RS, Carvalho HB, Massad E, Menezes RX, et al
. HIV infection and related risk behaviours in a disadvantaged youth institution of Sao Paulo, Brazil. Int J STD AIDS
13.Stricof RL, Kennedy JT, Nattell TC, Weisfuse IB, Novick LF. HIV seroprevalence in a facility for runaway and homeless adolescents. Am J Public Health
14.Sweeney P, Lindegren ML, Buehler JW, Onorato IM, Janssen RS. Teenagers at risk of human immunodeficiency virus type 1 infection. Results from seroprevalence surveys in the United States. Arch Pediatr Adolesc Med
15.Noell J, Rohde P, Ochs L, Yovanoff P, Alter MJ, Schmid S, et al
. Incidence and prevalence of chlamydia, herpes, and viral hepatitis in a homeless adolescent population. Sex Transm Dis
16.Roy É, Haley N, Leclerc P, Cédras L, Blais L, Boivin J-F. Drug injection among street youth in Montreal: predictors of initiation. J Urban Health
17.Remis RS, Alary M, Otis J, Mâsse B, Demers E, Vincelette J, et al
. No increase in HIV incidence observed in a cohort of men who have sex with other men in Montreal. AIDS
18.Hogg RS, Weber AE, Chan K, Martindale S, Cook D, Miller ML, et al
. Increasing incidence of HIV infections among young gay and bisexual men in Vancouver. AIDS
19.Doherty MC, Garfein RS, Monterroso E, Brown D, Vlahov D. Correlates of HIV infection among young adult short-term injection drug users
20.Miller CL, Tyndall M, Spittal P, Li K, LaLiberte N, Schechter MT. HIV incidence and associated risk factors among young injection drug users. AIDS
21.Des Jarlais DC, Choopanya K, Millson P, Friedmann P, Friedman SR. The structure of stable seroprevalence HIV-1 epidemics among injecting drug users.
In: Drug injecting and HIV infection
. Edited by Stimson GV, Des Jarlais DC, Ball AL. Geneva: UCL Press for World Health Organization, 1998:91–100.
22.Hankins C, Alary M, Parent M, Blanchette C, Claessens C. The SurvUDI Working Group. Continuing HIV transmission among injection drug users in Eastern Central Canada: the SurvUDI Study 1995 to 2001. J Acquir Immune Defic Syndr
23.Batter V, Matela B, Nsuami M, Manzila T, Kamenga M, Behets F, et al
. High HIV-1 incidence in young women masked by stable overall seroprevalence among childbearing women in Kinshasa, Zaïre: estimating incidence from serial seroprevalence data. AIDS
24.Spittal PM, Craib KJP, Wood E, Laliberté N, Li K, Tyndall MW, et al
. Risk factors for elevated HIV incidence rates among female injection drug users in Vancouver. Can Med Assoc J
25.Strathdee SA, Galai N, Safaiean M, Celentano DD, Vlahov D, Johnson L, et al
. Sex differences in risk factors for HIV seroconversion among injection drug users. A 10 year perspective. Arch Intern Med
26.Fournier L, Chevalier S, Ostoj M, Caulet M, Courtemanche R, Plante N. Dénombrement de la Clientèle Itinérante dans les Centres d'Hébergement, les Soupes Populaires et les Centres de Jour des Villes de Montréal et Québec 1996–97, 2-Montréal: Premiers Résultats.
[Technical report] Santé-Québec, November 1998.