301449

International Epidemiological Association 2001 International Journal of Epidemiology 2001;30:1449–1454
Sex trade involvement and rates of human
immunodeficiency virus positivity among
young gay and bisexual men
Amy E Weber,a Kevin JP Craib,a Keith Chan,a Steve Martindale,a Mary Lou Miller,aMartin T Schechtera,b and Robert S Hogga,b Background Susceptibility to human immunodeficiency virus (HIV) infection is of particular concern for marginalized populations. The objective of this study was to deter-mine risk factors associated with sex trade work among young gay and bisexualmen. Further, we aimed to compare HIV prevalence and incidence among meninvolved and not involved in sex trade work.
The study is based upon data obtained from a prospective cohort study of younggay and bisexual men. Participants had completed a baseline questionnaire whichelicited information on demographic information, sexual behaviours, and sub-stance use. Sex trade involvement was defined as the exchange of money, drugs,goods, clothing, shelter or protection for sex within the one year prior to enrol-ment. Contingency table and multivariate logistic regression analyses were usedto identify risk factors associated with involvement in the sex trade.
Of the 761 eligible participants, 126 (16%) reported involvement in sex tradework. Multivariate logistic regression analysis revealed regular alcohol use (OddsRatio [OR] = 3.6, 95% CI : 1.8–7.2), aboriginal ethnicity (OR = 3.7, 95% CI : 1.6–8.7),unemployment (OR = 3.9, 95% CI : 2.1–7.3), history of residence in a psychiatricward (OR = 4.2, 95% CI : 1.8–9.8), bisexual activity (OR = 7.0, 95% CI : 3.5–14.1)and the use of crack (OR = 7.4, 95% CI : 3.0–18.7) to be independently associatedwith sex trade work. Sex trade workers had a significantly higher HIV prevalenceat baseline compared with non-sex trade workers (7.3% versus 1.1%, P Ͻ 0.001).
As well, HIV incidence was found to be significantly higher for sex trade workerscompared with non-sex trade workers (4.7% versus 0.9%, P = 0.011).
Our study reveals that for male sex trade workers in this setting increased vul-nerability to HIV infection is related to unfavourable living conditions, substanceuse and sexual risk behaviour.
Homosexual men, sex trade worker, risk factors, sexual behaviour, housing In Canada, gay and bisexual men have been seriously affected overall proportion of men infected with HIV, gay and bisexual by the human immunodeficiency virus/acquired immuno- men continue to be greatly affected by the HIV/AIDS epidemic.
deficiency symdrome (HIV/AIDS) epidemic. In 1985, 83% of In the US, gay and bisexual men accounted for 50% of AIDS all reported AIDS cases among adult males occurred in this cases and 43% of non-AIDS cases in 1996.2 population. By 1995, the percentage of reported cases in this Sexual transmission among gay and bisexual men continues category had decreased to 74%.1 Despite a reduction in the to be a major source of new HIV infections in Canada. As is truefor other populations, subgroups of men who have sex withmen may be at increased risk for HIV infection. Also, they may a British Columbia Centre for Excellence in HIV/AIDS, St Paul’s Hospital, be less likely to be reached by prevention messages aimed at bDepartments of Health Care and Epidemiology, University of British the broader gay and bisexual community.3 These vulnerable subgroups include men who inject drugs, trade sex for money Correspondence: Robert Hogg, Program Director, Division of Epidemiology or drugs and those men who do not self-identify as gay or and Population Health, BC Centre for Excellence in HIV/AIDS, 608–1081Burrard St, Vancouver, BC V6Z 1Y6, Canada. E-mail: [email protected] It is important to identify whether men who engage in the variables including the frequency of receptive and insertive anal sex trade are a demographically and behaviourally distinct sub- intercourse with regular and casual partners; and the frequency group of gay and bisexual men in order to develop better and of condom use during receptive and insertive anal intercourse.
more focused intervention programmes. This analysis aimed to The use of alcohol and other drugs was also assessed. For the determine risk factors associated with involvement in sex work purpose of this analysis unstable housing was defined as living among young gay and bisexual men in order to have a better in a hotel, boarding house, group home, in the street or having profile of their lives. Further, as men involved in sex work have been identified as being at increased risk of HIV infection, we The HIV prevalence and incidence calculations were also compared HIV prevalence and incidence for men involved and compared for the two groups. The HIV incidence was calculated as the number of new infections divided by the total person-time under observation. Person-time was calculated as theinterval between enrolment and the most recent follow-up visit for subjects who did not seroconvert through January 1999. For The Vanguard Project is an ongoing prospective study of over subjects who became HIV positive, person-time was calculated 900 gay and bisexual men in the Greater Vancouver region.
as the interval between enrolment and the first visit at which an Men were eligible to participate if they were aged 18–30 years, HIV positive test result was detected. The 95% CI for the incid- lived in the Greater Vancouver region, had not previously tested ence estimates were calculated based on the Poisson distribution.
HIV-positive, and self-identified as gay or bisexual or reported Categorical variables were compared between groups using having sex with other men. Potential participants were re- Pearson’s χ2 test. Contingency tables that contained one or more cruited through community outreach at gay community events, expected counts of less than five were analysed by the Fisher’s community health clinics or local physicians, and through the exact test. Comparisons of continuous variables were carried out using Wilcoxon’s rank-sum test. All variables that were foundto be significant at P р 0.05 in the univariate analyses were Study instrument
entered into a stepwise multivariate logistic regression model Since May 1995, participants completed a detailed self- that was used to identify independent risk factors associated administered questionnaire and provided a blood sample for HIV with involvement in the sex trade. Participants with missing antibody testing at baseline and annually thereafter. Completed data for variables entered into the multivariate model were baseline questionnaires provide demographic data as well as excluded from the analysis. All reported P-values are two-sided.
information regarding sexual behaviours. Included are aspectsof insertive and receptive anal and oral sex with regular partners (men with whom you have sex at least once per month), casualpartners (men with whom you have sex less than once per Of the 761 eligible participants, 126 (16%) reported that they month) and paid partners (sex exchanged for money, drugs, were involved in sex work. Comparison of sex trade workers goods, clothing, shelter or protection). Participants are also asked versus non-sex trade workers revealed significant differences whether they have ever been forced to have sex (any type of with respect to socio-demographic characteristics (Table 1). Sex sexual activity that you were forced or coerced into against your trade workers were younger (median: 23 years versus 26 years, will), and the age range when this first occurred. Participants P Ͻ 0.001), more likely to be Aboriginal (29% versus 6%, completed a seven-item abbreviated version of the Centres for P = 0.001), and to have less than a high school education Epidemiologic Studies of Depression (CES-D) scale. Participants (40% versus 12%, P Ͻ 0.001). Sex trade workers were more were classified as being clinically depressed if their score on the likely to report living in unstable housing (45% versus 4%, abbreviated CES-D scale was greater than the median value for P Ͻ 0.001), to have an annual income of less than $10 000 the cohort. Additionally, participants answered questions about (58% versus 26%, P Ͻ 0.001), and to have ever been in a psy- their use of legal and illegal recreational drugs including chiatric ward (32% versus 6%, P Ͻ 0.001). Further, sex trade alcohol, tobacco, nitrite inhalants (poppers), marijuana, cocaine workers were found to have a high depression score (82% and heroin. Baseline questionnaires were used to assess these versus 50%, P Ͻ 0.001) and to have ever been in jail (54% versus 6%, P Ͻ 0.001) compared with non-sex trade workers.
Table 2 outlines differences in reported sexual behaviours of Statistical analysis
sex trade workers and non-sex trade workers. Sex trade workers The analysis presented here is limited to those individuals who reported a younger median age at first consensual sex with both had completed a baseline questionnaire and HIV test between men (15 years versus 18 years, P Ͻ 0.001) and women (14 years May 1995 and January 1999. As noted above, sex trade in- versus 17 years, P Ͻ 0.001). Sex trade workers were found to volvement was defined as the exchange of money, drugs, goods, be significantly more likely to report bisexual activity (48% clothing, shelter or protection for sex. It is important to note versus 8%, P Ͻ 0.001), and to having ever experienced non- that for the purposes of this analysis any man who reported sex consensual sex (49% versus 32%, P Ͻ 0.001) compared with trade involvement in the year prior to baseline were included.
those men not involved in the sex trade.
In order to assess risk factors associated with involvement in the With respect to sexual encounters with regular and casual sex trade, we conducted cross-sectional comparative analyses.
partners, a lower proportion of sex trade workers reported un- Variables of interest in these comparative analyses included: protected insertive (50% versus 75%, P Ͻ 0.001) and receptive socio-demographic characteristics such as age, ethnicity, (53% versus 75%, P Ͻ 0.001) oral sex with regular partners.
income, housing status and education; sexual behaviour There were no statistical differences in the proportion of people DETERMINANTS OF INVOLVEMENT IN THE SEX TRADE Table 1 Reported socio-demographic characteristics of male sex trade workers and non-sex trade workers
Sex trade workers
Non sex trade workers
P-value
Ethnicity
Less than high school education
Unstable housing
Income Ͻ$10 000/year
Ever in psychiatric ward
High depression score (CES-D Ͼ12)
Ever in jail
NB: Column totals may not equal total number of subjects due to missing values.
Table 2 Reported sexual experiences of male sex workers and non-sex trade workers
Sex trade workers
Non sex trade workers
P-value
Age at first sexual experience with:
Bisexual activity in previous year
Non-consensual sex (ever)
NB: Column totals may not equal total number of subjects due to missing values.
reporting unprotected anal intercourse with regular partners or Substance use was also found to be higher among the sex unprotected oral sex with casual partners. However, higher levels trade workers compared with non-sex trade workers (Table 4).
of unprotected insertive (28% versus 15%, P = 0.002) and Compared with non-sex trade workers, sex trade workers receptive (20% versus 13%, P = 0.042) anal sex were reported were more likely to consume more than 10 drinks per week with casual partners for sex trade workers compared with non- (39% versus 13%). As well, a higher proportion of sex trade workers reported the use of poppers (43% versus 28%), cocaine Table 3 Reported unprotected sexual behaviour with regular and casual partners in the year prior to baseline for male sex trade workers and
non-sex trade workers
Sex trade workers
Non sex trade workers
P-value
Regular partners
Casual partners
NB: Column totals may not equal total number of subjects due to missing values.
Table 4 Comparison of substance use in the year prior to baseline for male sex trade workers and non-sex trade workers
Sex trade workers
Non sex trade workers
P-value
Greater than 10 drinks/week
Injected drugs
NB: Column totals may not equal total number of subjects due to missing values.
(63% versus 25%), crack (38% versus 4%), and heroin (30% 1.1%, P Ͻ 0.001). As well, the HIV incidence rate was found to versus 3%) compared with non-sex trade workers (all P Ͻ 0.001).
be significantly higher for sex trade workers 4.7/100 person years Sex trade workers were also more likely to report having (pyrs) (95% CI : 0.1–9.4/100 pyrs) compared with non-sex injected drugs in the year prior to baseline compared to non-sex trade workers 0.9/100 person years (95% CI : 0.3–1.5/100 pyrs) trade workers (32% versus 4%, P Ͻ 0.001).
The results of the logistic regression analysis of independent factors associated with involvement in the sex trade are sum- marized in Table 5. The regular use of alcohol characterized by having more than 10 drinks per week was associated with a Comparison of male sex trade workers and other young gay and 3.6-fold increase in the likelihood of being involved in sex trade bisexual men revealed many important differences with respect work. Further independent risk factors for involvement in sex to risk factors associated with sex work. Sex trade workers were trade work were aboriginal ethnicity (Odds Ratio [OR] = 3.7), more likely to be marginalized as characterized by their younger being unemployed (OR = 3.9), having ever been in a psychiatric age, low levels of education, unstable housing and low annual ward (OR = 4.2), engaging in sexual relations with both men incomes. Aboriginal status was found to be independently asso- and women (OR = 7.0) and using crack (OR = 7.4).
ciated with sex work. Although the proportion of aboriginal Sex trade workers had a significantly higher HIV prevalence men in our study who were involved in prostitution (20%) was at baseline compared with non-sex trade workers (7.3% versus much lower than that recently reported for youth in Canada,4 DETERMINANTS OF INVOLVEMENT IN THE SEX TRADE Table 5 Multivariate logistic regression model: Risk factors associated with involvement in the sex trade (n = 521)
Unadjusted odds ratio
Adjusted odds ratio
Use of alcohol
Aboriginal ethnicity
Unemployed
Psychiatric ward
Bisexual activity
Use of crack
Aboriginal men are over-represented in the sex trade worker Prostitution is an economic necessity for many drug users, population. To date there is little research of male sex trade both male and female.20 Our results indicate that a significantly workers from different ethnic backgrounds. However, Aboriginal higher proportion of sex trade workers used both non-injection men have been found to make up a significant proportion of and injection drugs compared with the rest of the cohort. The men involved in prostitution, street youth and injection drug proportion of sex trade workers reporting the use of poppers, users.5,6 Further evidence of the marginalized status of these cocaine, crack and heroin was significantly higher than other men include the higher proportion of sex trade workers report- gay and bisexual men in this study. People who use injection ing having ever been in a psychiatric ward, having a high depres- drugs are at particular risk for HIV infection because they may sion score and having ever been incarcerated. Unemployment be exposed to the virus through both unprotected sex and the and a history of residence in a psychiatric ward were found to sharing of injection equipment.21,22 de Graff et al.8 reported be independently associated with sex work. These differences that heavily addicted drug users were less discriminating in are consistent with other studies that have shown sex trade their choice of clients, and lowered their barriers more easily workers to be living in impoverished conditions confounded by when it came to performing sexual acts without using condoms.
mental illness and substance use.7–9
Substance use characterized by the consumption of more than Sex trade workers may be at increased risk of infection 10 drinks per week and the use of crack was found to be inde- with HIV due in part to their sexual and drug using behaviours.
pendently associated with sex work. The relationship between Men involved in sex work were on average younger at first substance use and sex work is complex. Studies have shown that consensual sex with both males and females compared with in some cases substance use precedes entry into prostitution non-sex trade workers. This finding corroborates earlier work while in others, sex work is used as a means to earn money by Earls and David10 that indicated that male prostitutes were to buy drugs.10,23 Regardless of the temporal relationship, it is significantly younger than the comparison group in terms of the important to recognise the relationship and acknowledge the age at first sexual interaction. Younger age at first coitus has importance of drug treatment for these men. The use of crack is been linked to risky sexual behaviour including multiple sexual of further concern because it has been linked with such sexual partners.11 Sex trade workers in our study were found to have behaviours as inconsistent condom use and multiple partners lower rates of unprotected oral sex with regular male partners that may increase an individual’s risk of infection with and higher rates of unprotected anal sex with casual male partners. The latter finding may contribute to the significantly The prevalence of HIV for male sex trade workers was signifi- higher HIV prevalence and incidence rates found for sex trade cantly higher than for other gay and bisexual men in this study.
workers compared with non-sex trade workers. Reporting of a However, the prevalence is fourfold lower than the rates reported history of non-consensual sex was higher for sex trade workers for male prostitutes in Atlanta, USA.28 The independent risk compared with the rest of the cohort. These data are consistent factors for sex trade involvement highlighted in this study may with other studies showing a relationship between non-consensual contribute to increased risk of HIV infection. Minority status sex and involvement in the sex trade.10,12–14 Bisexual activity and marginalization have been linked with increased HIV pre- was found to be independently associated with sex trade involve- valence.6,7,29 Involvement in bisexual activity may increase the ment. Previous research has suggested that men may self-identify risk of HIV infection due to the relationship between bisexual as heterosexual and report female sex partners but engage in activity and involvement in sex work.15 Substance use has been sexual activity with other men for profit.15–17 A number of studies linked with risk of HIV infection because the consumption of have suggested that male sex trade workers are a potential link alcohol and drugs may interfere with judgement and decision- between homosexual and heterosexual transmission of HIV.18,19 making thereby potentially increasing unsafe sexual behaviour.30 The danger of HIV infection spreading through homosexual As is the case with many population-based studies, this is a prostitution is very real, but the actual risk depends to a great sample of convenience. Prostitution is a highly heterogeneous extent on which sexual practices and precautionary measures activity, with sex being sold in saunas, bars, clubs, public toilets, through escort agencies and contact advertisements, as well as on the street.31 The sex trade workers involved in our study self- 10 Earls CM, David H. A psychosocial study of male prostitution. Arch Sex identified and were not asked the location from which they sell Behav 1989;18:401–19.
sex, and so it is difficult to say whether we have adequately 11 Greenberg J, Magder L, Aral S. Age at first coitus: A marker for risky represented the male sex worker population in Vancouver. As sexual behavior in women. Sex Transm Dis 1992;19:331–34.
well, the definition of sex work used in this study is very broad 12 Strathdee SA, Hogg RS, Martindale SL et al. Determinants of sexual and this must be taken into account when interpreting the risk-taking among young HIV-negative gay and bisexual men. J Acquir results. The issue of temporality must be taken into considera- Immune Defic Syndr Hum Retrovir 1998;19:61–66.
tion in the interpretation of the multivariate analysis. It is Bartholow BN, Doll LS, Joy D et al. Emotional behavioural and HIV important to note that we have not examined causation in the risks associated with sexual abuse among adult homosexual and
bisexual men. Child Abuse Neglect 1994;18:747–61.
analysis. Therefore it is not possible for us to determine whether the associated factors have a predictive or causative effect.
Zierler S, Feingold L, Laufer D, Velentgras P, Kantrowitz-Gordon I,Mayer K. Adult survivors of childhood sexual abuse and subsequent In summary, risk factors for involvement in the sex trade risk of HIV infection. Am J Public Health 1991;81:572–75.
were belonging to a minority, marginalization characterized by 15 Rietmeijer CA, Wolitski RJ, Fishbein M, Corby NH, Cohn DL. Sex unemployment and mental illness. Further, this study demon- hustling, injection drug use, and non-gay identification by men who strates that men involved in the sex trade are engaging in have sex with men. Associations with high-risk sexual behaviors and high-risk behaviours and their relatively high prevalence of HIV condom use. Sex Transm Dis 1998;25:353–60.
puts these men and their sexual partners at an increased risk of 16 Radford JL, King AJC, Warren WK. Street Uouth and AIDS. Kingston: transmission of HIV. This study highlights the need for person- Social Program Evaluation Group, Queen’s University, 1989.
alized prevention messages in conjunction with intensified 17 Pleak R, Meyer-Bahlburg H. Sexual behavior and AIDS knowledge of realistic and acceptable HIV prevention and intervention efforts.
young male prostitutes in Manhattan. J Sex Res 1990;27:557–87.
18 Boulton M, Hart G, Fitzpatrick R. The sexual behaviour of bisexual men in relation to HIV transmission. AIDS Care 1992;4:165–75.
19 Morse EV, Simon PM, Osofsky HJ, Balson PM, Gaumer HR. The male street prostitute: a vector for transmission of HIV infection into the This study is supported by a grant from the National Health heterosexual world. Soc Sci Med 1991;32:535–39.
Research and Development Programme (NHRDP), Health 20 van den Hoek JAR, Coutinho RA, van Haastrecht HJA, van Zadelhoff Canada. National Health Scholar Awards granted by the NHRDP, AW, Goudsmit J. Prevalence and risk factors of HIV infections among Health Canada, support Dr Hogg. Dr Schechter is an MRC drug users and drug using prostitutes in Amsterdam. AIDS scientist. The authors are indebted the participants, physicians, 1988;2:55–60.
nurses and clinic staff of the Vanguard Project cohort and 21 Lewis DK, Watters JK. Sexual behaviour and sexual identity in malethe Community Advisory Committee of the Vanguard Project.
injection drug users. J Acquir Immune Defic Syndr Hum Retrovir 1994;7:
Thank you to Peter Vann for his administrative assistance.
22 Mandell W, Vlahov D, Latkin C, Oziemkowska M, Cohn S. Correlates of needle sharing among injection drug users. Am J Public Health 1994;
84:902–23.
23 Silbert MH, Pines AM, Lynch T. Substance abuse and prostitution. 1 AIDS in Canada. Quarterly Surveillance Update. Division of HIV/AIDS J Psychoact Drugs 1982;14:193–97.
Surveillance Bureau of HIV/AIDS and STD Laboratory. Centre for 24 Fullilove AE, Thompson Fullilove M, Bowser BP, Gross SA. Risk of Disease Control Health Protection Branch, Health Canada, 1997.
sexually trnasmitted disease among black adolescent crack users in 2 Centres for Disease Control and Prevention. HIV/AIDS Surveillance Oakland and San Francisco, California. Sex Transm Dis 1990;263:851–55.
Report 1997;8:10–30.
25 Wilson TE, Minkoff H, DeHovitz J, Feldman J, Landesman S. The 3 Goldbaum G, Perdue T, Woliski R et al. Differences in risk behavior relationship of cocaine use and human immunodeficiency virus and sources of AIDS information among gay, bisexual, and straight- serostatus to incident sexually transmitted diseases among women. identified men who have sex with men. AIDS Behav 1998;2:13–21.
Sex Transm Dis 1998;25:70–75.
4 Save the Children Canada. More Aboriginal Youth Exploited in the Sex 26 Ross MW, Hwang L, Leonard L, Teng M, Duncan L. Sexual behaviour, STDs and drug use in a rack house population. Int J STD AIDS 1999; 5 deMatteo D, Major C, Block B et al. Toronto street youth and HIV/ 10:224–30.
AIDS: Prevalence, demographics and risks. J Adolesc Health 1999;25:
27 Siegal HA, Carlson RG, Falck R, Forney MA, Wang J, Li L. High-risk behaviors for transmission of syphilis and human immunodeficiency 6 Rekart ML, Chan S, Barnett J, Lawrence C, Manzon L. HIV and North virus among crack cocaine-using women: A case study from the American aboriginal peoples. International Conference on AIDS, Midwest. Sex Transm Dis 1992;19:266–71.
1991 June 16–21;7(1):357 Abstract #M.C. 3237.
28 Elifson KW, Boles J, Darrow WW, Sterk CE. HIV seroprevalence and 7 Tynes LL, Sautter FJ, McDermott BE, Winstead DK. Risk of HIV risk factors among clients of female and male prostitutes. J Acquir infection in the homeless and chronically mentally ill. Southern Med J Immune Defic Syndr Hum Retrovir 1999;20:195–200.
1993;86:276–81.
29 Susser E, Valencia E, Miller M, Tsai WY, Meyer-Bahlburg H, 8 de Graaf R, Vanwesenbeeck I, van Zessen G, Straver CJ, Visser JH.
Conover S. Sexual behavior of homeless mentally ill men at risk for Male prostitutes and safe sex: different settings, different risks. AIDS HIV. Am J Psychiatr 1995;152:583–87.
Care 1994;6:277–88.
30 de Graaf R, Vanwesenbeeck I, van Zessen G, Straver CJ, Visser JH.
9 Simon PM, Morse EV, Osofsky HJ, Balson PM, Gaumer HR.
Alcohol and drug use in heterosexual and homosexual prostitution, Psychological characteristics of a sample of male street prostitutes.
and its relation to protection behaviour. AIDS Care 1995;7:35–47.
Arch Sex Behav 1992;21:33–44.
31 Hart G, Whittaker D. Sex workers and HIV. AIDS Care 1994;6:267–68.

Source: http://www.momentumstudy.ca/sites/default/files/abstracts/vanguard/Sex_trade_involvement_and_rates.pdf