|
This
paper was published in the September 2001 issue of the Journal
of Acquired Immune Deficiency Syndromes (28,1:81-88).
Risk Factors Associated With HIV Infection Among Young
Gay and Bisexual Men in Canada
Amy E. Webera;
Keith Chana; Clemon Georgeb; Robert S. Hogga,c;
Robert S. Remisb,d,e; Steve Martindalea;
Joanne Otisb,f; Mary Lou Millera; Jean Vinceletteb,g;
Kevin J. P. Craiba; Benoit Mâsseb,h;
Martin T. Schechterc; Roger LeClercb,i;
René Lavoieb,j; Bruno Turmelb,e;
Raymond Parentb,e; and Michel Alaryb,g.
aBritish Columbia Centre for Excellence
in HIV/AIDS, St. Paul's Hospital, Vancouver;
bOmega Cohorte, Montreal;
cDepartment of Health Care and Epidemiology University
of British Columbia, Vancouver;
dDepartment of Public Health Sciences, University
of Toronto, Toronto;
eDépartement de santé publique, Montréal-Centre,
Montreal;
fDépartment de sexologie, Université
du Québec à Montréal, Montreal;
gCentre hospitalier de l'Université de Montréal,
Campus St. Luc, Montreal;
hGroupe de recherche en épidémiologie,
Centre hospitalier affilié universitaire, Pavillion St.-Sacrement,
Quebec;
iCOCQ-sida, Montreal; and
jSéro-zéro, Montreal, Canada.
View or download the PDF file of this paper.
ABSTRACT:
 Objective: To assess risk factors associated
with HIV prevalence and incidence among gay and bisexual men
in two prospective Canadian cohorts.
Methods: The
Vanguard Project and the Omega Cohort are prospective cohort
studies of gay and bisexual men ongoing in Vancouver and Montreal,
respectively. For this analysis, baseline sociodemographic characteristics,
sexual behavior, and substance use data from these two cohorts
were combined. Assessment of risk factors for HIV seroprevalence
and seroconversion were carried out using univariate and multivariate
analysis.
Results: This
analysis was based on 1373 gay and bisexual men aged 16 to 30
years. Men who were HIV-seropositive at baseline (n=48) were
more likely to report living in unstable housing, to have had
less than a high school education, and to have been unemployed
than those who were HIV-negative (n=1325). HIV-positive men were
also more likely to report having engaged in sexual risk behavior,
including having had consensual sex at a younger age, having
had at least 6 partners during the previous year, ever having
been involved in the sex trade, and having engaged in unprotected
receptive anal intercourse. With respect to substance use, HIV-positive
men were more likely to report the use of crack, cocaine, heroin,
and marijuana and to use injection drugs. Similarly, men who
seroconverted during the course of the studies (n=26) were more
likely to report having less than a high school education and
having lived in unstable housing at baseline. Compared with HIV-negative
men, men who seroconverted were more likely to report ever having
been involved in the sex trade and engaging in unprotected receptive
anal intercourse. Reports of cocaine use and injection drug use
were also significantly higher for men who seroconverted compared
with HIV-negative men.
Conclusions:
Our data indicate that HIV-positive gay and bisexual men are
more likely to be living in unstable conditions and to report
more risky sexual and substance use behaviors than HIV-negative
men.
Key Words: HIV/AIDS,
Sexual behavior, Gay men, Seroconversion.
Since the beginning of the HIV epidemic in North America, the
majority of HIV infections have occurred among men who engage
in sexual relations with other men. As the HIV epidemic enters
its third decade, gay and bisexual men continue to have among
the highest rates of HIV infection. Previous studies have highlighted
the decline in HIV incidence and risk behavior among gay and bisexual
men (1,2). Nevertheless, several recent studies
have suggested that young gay and bisexual men continue to engage
in unprotected sexual behaviors and are at continued risk of HIV
infection (3,4).
Numerous studies have examined risk factors for HIV infection
among men who have sex with men (5-7). Important risk factors for HIV
infection include marginalization (8,9); sex trade involvement (10,11); substance use, including the
use of cocaine (12-14);
and unprotected anal intercourse (14-17). Caceres and van Griensven (5) highlighted the facts that these studies
have experienced methodologic problems and that risk factor analyses
have produced contradictory results. Data on the specific predictors
of HIV infection among young gay and bisexual men are also lacking.
This study aimed to assess risk factors for prevalent HIV infection
and predictors of HIV seroconversion for men who have sex with
men from two prospective cohorts of gay and bisexual men. The
current study is of particular interest in that it focuses on
young gay and bisexual men, who continue to comprise a large proportion
of new HIV infections in North America.
METHODS
The Vanguard Project is a prospective study of more than 850
gay and bisexual men aged 15 to 30 years who live in the Greater
Vancouver region. These men are recruited through outreach, clinics,
and physicians' offices. To be eligible for this longitudinal
study, the participants must have not previously tested positive
for HIV and must self-identify as being gay or bisexual or having
sex with other men. Since May 1995, Vanguard Project participants
have completed a self-administered questionnaire and undergone
HIV antibody testing on an annual basis. The questionnaire elicits
information on sociodemographic characteristics; sexual behavior,
including involvement in the sex trade; and substance use in addition
to other psychosocial variables.
The Omega Cohort is similar to the Vanguard Project in that
it is a prospective study of over 1400 HIV-negative men who self-identify
as being gay or bisexual or having sex with other men. The Omega
Cohort recruits men through outreach and publicity as well as
through medical clinics and physicians. The minimum age for recruitment
in the Omega Cohort is 16 years of age, and there is no maximum
age limit. Follow-up procedures for the Omega Cohort consist of
semiannual interviews, self-administered questionnaires, and HIV
antibody testing.
To assess factors associated with HIV-positive
serostatus,
we combined the data from the two cohorts and conducted cross-sectional
comparative analyses. To ensure comparability between the two
cohorts, the age at enrollment was restricted to 16 to 30 years
for participants in the Omega Cohort. There are currently no Vanguard
Project participants younger than 16 years of age; thus, we were
able to match the age of Vanguard Project subjects with that of
Omega Cohort subjects. There were 770 eligible men from the Vanguard
Project and 603 from the Omega Cohort. Seven hundred thirty-six
(56%) of the HIV-negative men were participants in the Vanguard
Project, with the remainder (44%) being participants in the Omega
Cohort. Of the 48 men who were HIV prevalent, 34 (71%) were participants
in the Vanguard Project and 14 (29%) were from the Omega Cohort.
We also examined baseline predictors of HIV seroconversion. Over
the course of the respective study periods, 26 men seroconverted:
18 (69%) from the Vanguard Project and 8 (31%) from the Omega
Cohort.
Variables of interest in these analyses included sociodemographic
characteristics such as age, ethnicity, income, housing status,
and education; sexual experiences such as age at first consensual
sex, number of sexual partners during the previous 6 months or
1 year, ever having been paid for sex, and occurrences of unprotected
insertive and receptive anal intercourse as well as substance
use. For this analysis, unstable housing was defined as living
in a hotel, boarding house, group home, or on the street or having
no fixed address.
The time frame of the questions in the two cohorts was different.
The Vanguard Project refers to situations that may have occurred
ever or during the year before baseline. The Omega Cohort uses
the 6 months before baseline as the reference period for some
of the questions and lifetime for others. Whenever possible, we
attempted to match the time frames for given questions; however,
because of the lack of subsequent follow-up data for the seroprevalent
men in the Omega Cohort, the reference time periods for the two
groups may differ by 6 months for certain questions. A further
data-related consideration was the difference in data collection
methods employed by the two cohorts. Despite the fact that the
Omega Cohort used both interviewer-administered and self-administered
questionnaires, all but two pieces of data (date of birth and
age of first consensual sex) were obtained through self-administered
questionnaires.
Statistical Methods
Categoric variables were compared for the two groups using
the Pearson X2 test. Contingency tables that contained
one or more expected counts of <5 were analyzed by the Fisher
exact test. Comparisons of continuous variables were carried out
using the Wilcoxon rank sum test. All variables that were significant
in the univariate analysis were made available for stepwise multivariate
logistic regression. Multivariate logistic regression analyses
were used to identify independent baseline factors associated
with HIV-positive serostatus and HIV seroconversion. Multivariate
logistic regression analyses were restricted to men reporting
anal receptive intercourse during the 6 months to 1 year before
baseline to assess the relation between condom use for receptive
anal sex and HIV-positive serostatus. All probability values are
two-sided.
RESULTS
Risk Factors for Baseline HIV-Positive Serostatus
There were 1373 gay and bisexual men between the ages of 16
and 30 years (770 from the Vanguard Project and 603 from Omega
Cohort) who had completed a baseline questionnaire and HIV antibody
test. Comparison of sociodemographic characteristics of HIV-negative
men and men who were HIV-positive at baseline revealed no significant
differences with respect to age, aboriginal status, or having
a low income level (Table
1). HIV-positive men were more likely to report living
in unstable housing (19% vs. 6%; p<.002) compared with HIV-negative
men. Compared with HIV-negative men, HIV-positive men were also
more likely to report having completed less than a high school
education (34% vs. 12%; p<.001) and being unemployed at the
baseline visit (43% vs. 27%; p=.016).
Table
2 outlines baseline sexual behavior of HIV-negative men
compared with men who tested HIV-positive. Men who were HIV-positive
were significantly more likely to report being younger at their
first consensual sexual experience (median: 15 years vs. 17 years;
p=.002). HIV-positive men were also significantly more likely
to have had 6 or more sexual partners during the previous 6 months
or 1 year (65% vs. 49%; p=.028), to have ever been involved in
the sex trade (51% vs. 25%; p<.001), and to have engaged in
unprotected receptive anal intercourse during the previous 6 months
or 1 year (56% vs. 40%; p=.035). HIV-positive men were also more
likely to report having engaged in unprotected sex outside their
province of residence and to have engaged in unprotected insertive
anal intercourse, although these differences were not statistically
significant.
Comparison of reported substance use revealed no significant
differences between the two groups with respect to the use of
nitrite inhalants ("poppers") or acid (Table 3). HIV-positive men were significantly
more likely to report ever having used crack (17% vs. 5%; p<.003),
cocaine (58% vs. 30%; p<.001), heroin (15% vs. 6%; p=.019),
and marijuana (80% vs. 65%; p=.032). With respect to injection
drug use, men with a positive HIV test result were also more likely
to report having ever used injection drugs (26% vs. 7%; p<.001).
Multivariate logistic regression analysis was used to assess
risk factors associated with HIV-positive serostatus (Table 4). For all men, having less
than a high school education, history of cocaine use, and having
ever been involved in the sex trade were independently associated
with HIV-positive serostatus. We also examined men who reported
engaging in receptive anal intercourse to determine if unprotected
anal sex was associated with HIV-positive serostatus. Only two
factors, history of cocaine use and having ever been involved
in the sex trade, were found to be independently associated with
HIV prevalence.
Predictors of Seroconversion
Comparative analyses of baseline variables of HIV-negative
men (n=1325) and men who were HIV-negative at baseline and subsequently
seroconverted (n=26) were conducted (Table 5). Comparison of sociodemographic
variables revealed that men who seroconverted were significantly
more likely to report having less than a high school education
(27% vs. 12%; p=.027) and to have been living in unstable housing
at baseline (19% vs. 6%; p=.020). There were no significant differences
between the two groups with respect to baseline age, aboriginal
status, having a low income, or being unemployed (data not shown).
Men who seroconverted were of similar age to HIV-negative men
at their first consensual sexual experience. Similar proportions
of the two groups reported having had at least 5 sexual partners
during the previous 6 months or 1 year. Men who seroconverted
were more likely to report ever having been involved in the sex
trade (50% vs. 25%; p=.003) and having engaged in unprotected
receptive anal intercourse during the 6 months to 1 year before
baseline (60% vs. 40%; p=.041), however. Although not statistically
significant, men who seroconverted were also more likely to report
ever engaging in unprotected sex outside their province of residence
(44% vs. 27%; p=.068).
With respect to substance use, only reported history of cocaine
use was significantly different between the two groups, with 54%
of seroconverters reporting ever using cocaine compared with 30%
of HIV-negative men (p=.009). Further, men who seroconverted were
significantly more likely to report having ever injected drugs
(19%) compared with HIV-negative men (7%) (p=.031).
After adjusting for cohort, analysis of baseline predictors
of seroconversion revealed that for all men, having ever been
involved in the sex trade was associated with a threefold increase
in the odds of seroconversion (odds ratio [OR]=3.08, 95% confidence
interval [CI]: 1.41-6.73). Among men reporting receptive anal
intercourse, living in unstable housing (OR=4.26, CI: 1.48-12.29)
and unprotected receptive anal intercourse (OR=2.35, CI: 1.04-5.30)
were found to be predictive of HIV seroconversion.
Subanalyses
To ensure comparability of the two cohorts, the univariate
analysis was stratified by cohort (Vanguard Project vs. Omega
Cohort). With respect to the comparison of seroprevalent and seronegative
men, the trends were the same for the men from both cohorts for
all variables with the exceptions of the use of crack (for Vanguard
Project, OR=3.21, CI: 1.39-7.39; for Omega Cohort, no crack use),
having unprotected receptive anal intercourse (for Vanguard Project,
OR=2.41, CI: 1.18-4.93; for Omega Cohort, OR=0.96, CI: 0.27-3.46),
and having had at least 6 partners during the previous 6 months
(for Vanguard Project, OR=2.52, CI: 1.16-5.50; for Omega Cohort,
OR=1.03, CI: 0.34-3.09). Comparison of HIV-negative men and those
men who seroconverted in the two cohorts revealed trends similar
to the combined results for all variables with the exception of
unprotected receptive anal intercourse at baseline (for Vanguard
Project, OR=3.13, CI: 1.16-8.46; for Omega Cohort, OR=1.08, CI:
0.24-4.90). Two-way interactions between the independent variables
and the cohort variable were examined in the multivariate analyses.
No significant interactions were detected (p<.15).
As a further subanalysis, men who reported a history of injection
drug use were excluded so as to explore sexual transmission of
HIV among young gay and bisexual men. Thus, 142 and 41 men who
reported a history of injection drug use were excluded from the
seroprevalence and seroconversion analyses, respectively. Examination
of sociodemographic variables for HIV-positive and HIV-negative
men revealed that living in unstable housing (11% vs. 3%; p=.036)
was significantly associated with HIV prevalence. With respect
to sexual behavior, HIV-positive men were found to report a significantly
younger age at their first consensual sexual encounter (16 years
vs. 17 years; p=.024). Forty percent of HIV-positive men reported
ever having been involved in the sex trade compared with 21% of
HIV-negative men (p=.009). Several differences were found between
HIV-positive and HIV-negative men with respect to the use of drugs,
including poppers (40% vs. 25%; p=.049), cocaine (44% vs. 26%;
p=.017), and marijuana (80% vs. 64%; p=.048). Comparison of HIV-negative
men and men who seroconverted revealed significant sexual behavior
differences for unprotected sex outside the province of residence
(48% vs. 27%; p=.047) and having ever been paid for sex (43% vs.
21%; p=.030). There were no other significant differences found
between the two groups.
DISCUSSION
HIV Prevalence
The results of this study confirm many of the previously suggested
risk factors for HIV infection. Among the sociodemographic variables
examined, a marginalized lifestyle characterized by a low level
of education, unstable housing, and unemployment was associated
with an elevated risk of prevalent HIV infection. Several sexual
risk behaviors, including younger age at initiation into sexual
behavior, higher numbers of recent sexual partners, a history
of involvement in the sex trade, and engagement in unprotected
receptive anal intercourse, were identified for men who were HIV-positive
at baseline. Substance use, particularly the use of crack, cocaine,
heroin, and marijuana, was found to be associated with HIV-positive
prevalence. HIV-positive men were also more likely to report having
ever injected drugs.
Multivariate analysis of risk factors associated with baseline
seroprevalence revealed low education to be an independent risk
factor. This finding supports the results of studies by Osmond
et al. (8) and Ruiz et al. (9),
both of whom found a high level of education to be inversely associated
with HIV prevalence. Level of education may be a surrogate marker
for lifestyle stability in that persons with higher levels of
education may lead more stable lives and thereby be less inclined
to engage in behaviors that put them at increased risk of HIV.
Involvement in the sex trade was also identified as an independent
risk factor for HIV-positive serostatus. Several studies have
shown that working in the sex trade increases the risk of HIV
infection (10,11). The high number
of sexual partners and the potential for unprotected sexual behavior
in exchange for money may substantially increase the risk of HIV
transmission. Lifetime history of cocaine use was found to be
independently associated with HIV prevalence. The relation between
cocaine use and HIV prevalence may have two possible explanations.
First, the use of drugs, including cocaine, has been associated
with unprotected sex, thereby increasing the possibility of HIV
infection. Second, cocaine is often administered through a syringe,
and injection drug use is a known risk factor for HIV. Injection
drug use was found to be significant in univariate analysis but
was not found to be an independent risk factor for HIV prevalence.
Among men reporting having engaged in anal receptive intercourse
in the previous year, use of cocaine and having ever been involved
in the sex trade were found to be independently associated with
HIV prevalence. As stated previously, use of cocaine may increase
the risk of HIV infection by leading to a decrease in protective
behaviors. Cocaine has been shown to be one of the most important
predictors of unsafe sex or seroconversion for gay and bisexual
men (12-14). Additionally, the biologic
interaction model suggests that during unprotected anal receptive
intercourse, certain drugs, especially cocaine and poppers, increase
the number of physical pathways for HIV infection by relaxing
the anal sphincter and triggering vasodilatation (13).
Men involved in the sex trade may be more likely to engage in
anal sex because of the financial incentive, thereby increasing
their risk of HIV infection. There may be potential for greater
financial gain by engaging in unprotected sex, increasing the
risk of HIV infection.
HIV Incidence
Many previous studies have focused on risk factors for HIV
prevalence (4,6-9,18).
As much as these analyses provide information that is useful to
identify associations, we believed it was also important to examine
risk factors associated with seroincidence in an attempt to investigate
more causative relations. Evaluation of sociodemographic variables
revealed that a marginalized lifestyle characterized by low education
and unstable housing was associated with HIV seroconversion. Most
seroconverting men reported risky sexual behavior, in particular,
having ever been involved in the sex trade and engaging in unprotected
anal receptive intercourse during the year before baseline. Lifetime
use of cocaine and injection drugs was also found to be reported
by a significantly higher proportion of men who seroconverted
compared with HIV-negative men.
Multivariate analysis of baseline predictors of HIV seroconversion
revealed involvement in the sex trade to be an independent predictor
of seroconversion. Male prostitutes often engage in identified
risk behaviors, including multiple partners and engaging in activities
that place them at high risk for HIV infection (19). Men who are involved in the sex
trade also frequently engage in injection drug use, which puts
them at increased risk of HIV seroconversion. Among men who reported
engaging in anal receptive intercourse during the year before
their baseline visit, unstable housing was found to be an independent
predictor of HIV seroconversion. Homelessness has been recognized
in the United States and elsewhere as an important public health
concern. Street-involved individuals, particularly youth, may
engage in behaviors that put them at an elevated risk of HIV infection
and transmission (20-22).
Unprotected anal intercourse, particularly receptive anal intercourse,
has long been identified as an important route of transmission
for HIV. In this analysis, unprotected receptive anal sex was
found to be an independent predictor of HIV seroconversion. This
finding is corroborated by many other epidemiologic studies (14-17,23). Despite widespread dissemination
of information regarding the risk of HIV transmission through
unprotected sex, young gay and bisexual men continue to engage
in high-risk sexual behavior. Among our sample, 40% of men who
were HIV-negative at the time of this study reported engaging
in unprotected receptive anal intercourse. It is important, however,
to note that the type of partnership between men was not accounted
for in this analysis. As reported by Alary et al. (24), the odds of seroconversion for men
who reported unprotected anal sex with a regular HIV-negative
partner was 0.4, whereas the odds of seroconversion for any other
partner whether regular or casual was 7.8. These findings highlight
the importance of careful definition of unsafe sex among gay and
bisexual men.
As is the case with many population-based studies, this is
a sample of convenience. Thus, our sample may not be representative
of the general population of gay and bisexual men. Further, the
inclusion criteria stipulate that men have not previously tested
HIV-positive; thus, the population, particularly for the seroprevalence
analysis, is likely not representative of gay and bisexual men
in general. A second limitation of this study is the small number
of men who seroconverted. The small number may have resulted in
a decreased power to detect other predictors of seroconversion
in this population. In this study, we combined data from two cohort
studies to examine risk factors for HIV-positive serostatus. As
with any study of this nature, there are limitations. There were
particular questions, for example, those relating to oral sex,
that could not be combined because of differences in the manner
in which the data were collected. Thus, there were potential risk
factors that were not investigated in this study. The time frame
for some questions also differed between the two cohorts. For
example, one cohort inquired about a given activity during the
previous year, although the other asked about that activity during
an individual's lifetime. The issue of temporality must be considered
in the interpretation of the multivariate analysis of the seroprevalence
data.
Further analysis of risk factors for seroconversion may provide
a better idea of important causal risk factors for HIV infection
among young gay and bisexual men. For the purpose of this analysis,
however, baseline predictors were used as a proxy for behaviors
that may have occurred just before the time of seroconversion.
We believe that the use of baseline variables as predictors of
seroconversion in this data set was appropriate to limit the effect
of learning associated with participation in a cohort study. Because
data from two different studies were combined, there was no way
to control for the effect of cohort participation (i.e., changes
of responses over time as a result of cohort participation and
the effect of prevention counseling), which may differ between
the two studies. By using data from the baseline visit, the effect
of participation and learning was minimized. Further, analysis
of changing risk behavior over time for the two cohorts indicated
no significant change over the course of follow-up. Thus, we believe
that baseline data were a fair proxy for later behavior.
In summary, this study supports previous research findings
as to the risk factors associated with HIV infection. Despite
many years of intervention programs targeted at the gay community,
young gay and bisexual men continue to engage in behaviors that
put them at risk for HIV infection. Innovative HIV prevention
campaigns are necessary to reduce the spread of this disease in
at-risk populations.
TABLE 1.
Comparison of sociodemographic characteristics of HIV-positive
and HIV-negative gay and bisexual men at baseline
|
HIV-negative
n=1325*
n (%) |
HIV-positive
n=48*
n (%) |
p-value |
| Age (in years) |
| Median (IQR) |
25 (2228) |
25 (2228) |
.992 |
| Aboriginal |
| Yes |
73 (6) |
5 (11) |
.204 |
| No |
1124 (94) |
41 (89) |
| Unstable housing |
| Yes |
79 (6) |
9 (19) |
.002 |
| No |
1228 (94) |
38 (81) |
| Less than high school education |
| Yes |
150 (12) |
16 (34) |
<.001 |
| No |
1148 (88) |
31 (66) |
| Income <$10,000 per year |
| Yes |
451 (37) |
12 (29) |
.272 |
| No |
772 (63) |
30 (71) |
| Employed |
| Yes |
953 (73) |
27 (57) |
.016 |
| No |
345 (27) |
20 (43) |
|
* Sum
total is not equal to total because of missing values.
IQR: interquartile range. |
[return to
text]
TABLE 2.
Comparison of sexual behaviors for HIV-positive and HIV-negative
gay and bisexual men at baseline
|
HIV-negative
n=1325*
n (%) |
HIV-positive
n=48*
n (%) |
p-value |
| Age at first consensual sex |
| Median (IQR) |
17 (1420) |
15 (1318) |
.002 |
| Number of partners (previous year or 6 months) |
| 05 |
630 (51) |
30 (35) |
.028 |
| 6 or more |
662 (49) |
16 (65) |
| Unprotected sex outside the province |
| Yes |
317 (27) |
16 (37) |
.162 |
| No |
837 (73) |
27 (63) |
| Ever involved in sex trade |
| Yes |
320 (25) |
24 (51) |
.001 |
| No |
983 (75) |
23 (49) |
| Unprotected anal insertive intercourse |
| Yes |
455 (41) |
21 (49) |
.315 |
| No |
651 (59) |
22 (51) |
| Unprotected anal receptive intercourse |
| Yes |
438 (40) |
24 (56) |
.035 |
| No |
664 (60) |
19 (44) |
|
*
Sum total is not equal to total due to missing values.
IQR: interquartile range. |
[return to
text]
TABLE 3.
Comparison of substance use for HIV-negative and HIV-positive
gay and bisexual men at baseline
|
HIV-negative
n=1325*
n (%) |
HIV-positive
n=48*
n (%) |
p-value |
| Poppers |
| Yes |
331 (26) |
16 (35) |
.169 |
| No |
956 (74) |
30 (65) |
| Crack |
| Yes |
65 (5) |
8 (17) |
<.003 |
| No |
1220 (95) |
38 (83) |
| Acid |
| Yes |
253 (20) |
13 (28) |
.151 |
| No |
1034 (80) |
33 (72) |
| Cocaine |
| Yes |
387 (30) |
26 (58) |
<.001 |
| No |
907 (70) |
19 (42) |
| Heroin |
| Yes |
73 (6) |
7 (15) |
.019 |
| No |
1214 (94) |
40 (85) |
| Marijuana |
| Yes |
845 (65) |
37 (80) |
.032 |
| No |
452 (35) |
9 (20) |
| Injection drug use |
| Yes |
88 (7) |
12 (26) |
<.001 |
| No |
1212 (93) |
35 (74) |
|
*
Sum total is not equal to total due to missing values. |
[return to
text]
TABLE 4.
Multivariate logistic regression models: risk factors associated
with baseline HIV-positive serostatus for all men (n=1373) and
men who reported anal receptive intercourse (n=1145), adjusted
for cohort
|
Adjusted odds ratio |
95% confidence interval |
| Model 1: all men |
| Cocaine use |
2.23 |
1.144.35 |
| Less than a high school education |
2.15 |
1.064.41 |
| Ever involved in sex trade |
2.05 |
1.054.00 |
| Model 2: men reporting engaging in anal receptive
intercourse |
| Ever involved in sex trade |
2.49 |
1.264.93 |
| Cocaine use |
2.31 |
1.164.58 |
[return to
text]
TABLE 5.
Comparison of sociodemographic, sexual, and substance use behavior
for HIV-negative and HIV-seroconverted gay and bisexual men
|
HIV-negative
n=1325*
n (%) |
HIV-positive
n=48*
n (%) |
p-value |
| DEMOGRAPHICS: |
| Less than a high school education |
| Yes |
150 (12) |
7 (27) |
.027 |
| No |
1148 (88) |
19 (73) |
| Unstable housing |
| Yes |
79 (6) |
5 (19) |
.020 |
| No |
1228 (94) |
21 (81) |
| SEXUAL BEHAVIOR: |
| Ever involved in sex trade |
| Yes |
320 (25) |
13 (50) |
.003 |
| No |
983 (75) |
13 (50) |
| Unprotected anal receptive intercourse |
| Yes |
438 (40) |
15 (60) |
.041 |
| No |
664 (60) |
10 (40) |
| SUBSTANCE USE: |
| Cocaine |
| Yes |
387 (30) |
14 (54) |
.009 |
| No |
907 (70) |
12 (46) |
| Injection drug use |
| Yes |
88 (7) |
5 (19) |
.031 |
| No |
1212 (93) |
21 (81) |
|
*
Sum total is not equal to total due to missing values. |
[return to
text]
Acknowledgments: This study was supported by a grant
from the National Health Research and Development Program (NHRDP),
Health Canada. R.S. Hogg is supported by an Investigator Award
granted by the Canadian Institutes of Health Research. M. Alary
is the recipient of a Research Scholar Award from Funds de la
Recherche en Santé du Quebec (970097).
REFERENCES
1. van Griensven GJP, De Vroome
EMM, et al. Changes
in sexual behaviour and the fall in incidence of HIV infection
among homosexual men. BMJ 1989;298:218-21.
[return to text]
2. Kingsley LA,
Zhou SYJ, Bacellar H, et al. Temporal trends in human immunodeficiency
virus type 1 seroconversion 1984-1989: a report from the Multicenter
AIDS Cohort Study (MACS). Am J Epidemiol 1991;134:331-9.
[return to text]
3. Strathdee SA, Hogg RS, Martindale SL, et al. Determinants of
sexual risk-taking among young HIV-negative gay and bisexual men.
J Acquir Immune Defic Syndr Hum Retrovirol 1998;19:61-6.
[return to text]
4. Hays RB, Kegeles SM, Coates TJ. High HIV risk-taking among
young gay men. AIDS 1990;4:901-7.
[return to text]
5. Caceres CF, van Griensven
GJP. Male homosexual transmission of HIV-1. AIDS 1994;8:1051-61.
[return to text]
6. Lemp GF, Hirozawa AM, Givertz D, et al. Seroprevalence
of HIV and risk behaviors among young homosexual and bisexual
men: the San Francisco/Berkeley Young Men's Survey. JAMA 1994;
272:449-54.
[return to text]
7. Seage GRI, Mayer KH, Lenderking WR, et al. HIV and hepatitis
B infection and risk behavior in young gay and bisexual men. Public
Health Rep 1997;112:158-67.
[return to text]
8. Osmond DH, Page K, Wiley
J, et al. HIV infection in homosexual and bisexual men 18-29 years
of age: The San Francisco Young Men's Health Study. Am J Public
Health 1994;84:1933-7.
[return to text]
9. Ruiz J, Facer M, Sun RK. Risk factors for human immunodeficiency
virus infection and unprotected anal intercourse among young men
who have sex with men. Sex Transm Dis 1998;25: 100-7.
[return to text]
10. Wolitski RJ, Fishbein
M, Johnson WD, et al. Sources of HIV information among injecting
drug users: association with gender, ethnicity, and risk behaviour.
AIDS Community Demonstration Projects. AIDS Care 1996;8:541-55.
[return to text]
11. Rietmeijer CA, Wolitski RJ, Fishbein M, et al. Sex
hustling, injection drug use, and non-gay identification by men
who have sex with men. Associations with high-risk sexual behaviors
and condom use. Sex Transm Dis 1998;25:353-60.
[return to text]
12. Burcham JL, Tindall
B, Marmor M, et al. Incidence and risk factors for human immunodeficiency
virus seroconversion in a cohort of Sydney homosexual men. Med
J Aust1989;150:634-9.
[return to text]
13. Ostrow DG, DiFranceisco WJ, Chmiel JS, et
al. A case-control study of human immunodeficiency virus Type
1 seroconversion and risk-related behaviours in the Chicago MACS/CCS
cohort, 1984-1992. Am J Epidemiol 1995;142:875-83.
[return to text]
14. van Griensven GJP, Tielman
RAP, Goudsmit J, et al. Risk factors and prevalence of HIV antibodies
among homosexual men in the Netherlands. Am J Epidemiol 1987;125:1048-57.
[return to text]
15. Schechter MT, Boyko WJ, Douglas B, et al. The Vancouver
Lymphadenopathy/AIDS Study. HIV seroconversion in a cohort of
homosexual men. Can Med Assoc J 1986;135:1355-60.
[return to text]
16. Moss AR, Osmond DH, Cherman JC, et al. Risk factors
for AIDS and HIV seropositivity in homosexual men. Am J Epidemiol
1987; 125:1035-47.
[return to text]
17. Darrow WW, Echenberg DF, Jaffe HW, et al. Risk factors
for HIV INFECTION AMONG YOUNG GAY AND BISEXUAL MEN IN human immunodeficiency
virus (HIV) infections in homosexual men. Am J Public Health 1987;77:479-83.
[return to text]
18. Hart GJ, Flowers P, Der GJ, et al. Homosexual men's HIV
related sexual risk behaviour in Scotland. Sex Transm Infect 1999;75:
242-6.
[return to text]
19. Pleak R, Meyer-Bahlburg H. Sexual behavior
and AIDS knowledge of young male prostitutes in Manhattan. J Sex
Res 1990;27: 557-87.
[return to text]
20. Clatts MC, Rees Davis W, et al. Correlates
and distribution of HIV risk behaviors among homeless youths in
New York City: implications for prevention and policy. Child Welfare
1998;77:195-207.
[return to text]
21. Clatts MC, Davis WR. A demographic and behavioral profile
of homeless youth in New York City: implications for AIDS outreach
and prevention. Med Anthropol Q 1999;13:365-74.
[return to text]
22. Rotheram-Borus M, Kooman C, Ehrhardt A. Homeless youth
and HIV infection in the United States. Am Psychol 1991;46:1188-97.
[return to text]
23. Kuiken CL, van Griensven GJP, de Vroome
EMM,
et al. Risk factors and changes in sexual behavior in male homosexuals
who seroconverted for human immunodeficiency antibodies. Am J
Epidemiol 1990;132:523-30.
[return to text]
24. Alary M, Remis RS, Otis J, et al. HIV incidence among homosexual
and bisexual men in the Montreal Omega Cohort Study: unprotected
anal sex remains the main risk factor if properly defined [abstract].
Can J Infect Dis 2000;11:58B.
[return to text]
Manuscript received October 13, 2000; accepted June
1, 2001.
© 2001 Lippincott Williams & Wilkins, Inc., Philadelphia
For more information, contact:
Bonnie Devlin
Vanguard Project Coordinator
608 - 1081 Burrard Street
Vancouver, BC, Canada, V6Z 1Y6
Tel: (604)806-8306
Fax: (604)806-9044
|
