 

 

m LIBRARY
l Michigan State
Zuni University

 

 

 

This is to certify that the
thesis entitled

. .-t-.-,—.-

Risk and Protective Factors Related to HIV-risk Behavior: A
Comparison Between HIV-positive and HIV-negative Men Who
Have Sex With Men

presented by
Jason C. Fomey

has been accepted towards fulﬁllment I
of the requirements for the

degree in /P3ychology

ﬂ %ssor’ s Signature

6/15 20%?

Date

 

 

 

MSU is an afﬁrmative-action. equal-opportunity employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K;lProj/Aoc&Pres/CIRC/Date0ue mdd

RISK AND PROTECTIVE FACTORS RELATED TO HIV-RISK BEHAVIOR: A
COMPARISON BETWEEN HIV-POSITIVE AND HIV-NEGATIVE YOUNG MEN
WHO HAVE SEX WITH MEN
By

Jason C. Fomey

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF ARTS
Department of Psychology

2008

ABSTRACT
RISK AND PROTECTIVE FACTORS RELATED TO HIV-RISK BEHAVIOR: A
COMPARISON BETWEEN HIV-POSITIVE AND HIV-NEGATIVE YOUNG MEN
WHO HAVE SEX WITH MEN
By
Jason C. F omey

The objective of this study was to assess and compare the prevalence of high risk sexual
behaviors among young HIV-negative (n = 8,064) and HIV-positive (n = 171) men who
have sex with men (MSM) on predictors of unprotected anal intercourse (UAI). Using
venue-based time-space sampling, 8,235 MSM aged 15-25 were anonymously surveyed
as a part of the Community Intervention Trial for Youth (CITY). The Project was
conducted in 13 communities across the United States from 1999-2002. As hypothesized,
there were few differences between the serostatus group rates of risky sexual behaviors,
which were high for all of the sexually active men. It was found that 40% of HIV-
positive men and 34% of HIV-negative men reported to have had UAI in the previous 3
months. HIV-positive MSM were more likely to have traded sex for money, food, drugs
or a place to stay within the previous year and to have had sex while high during their last
sexual encounter than were their uninfected peers. HIV-infected men also had UAI with
more sexual partners than did those who were uninfected. Multivariate analyses indicated
that for HIV-negative men, positive peer norms regarding safer sex and being Black or
Latino predicted avoidance of UAI. Among HIV-positive men, having social support for
safer sex and positive peer norms predicted avoidance of UAI. Our ﬁndings suggest that

young HIV-positive MSM are a relevant subgroup for prevention because they constitute

a signiﬁcant source from which future infections could be generated.

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor Dr. Robin Miller for her
encouragement, mentorship, and critical feedback throughout the writing of this thesis. I
am also enormously grateful to my supportive committee, Dr. Cris Sullivan and Dr. Zaje
Harrell. Additionally, many thanks to my wonderful colleagues in the Ecological-
Community Psychology program for their support and feedback. I am also indebted to
my good friend Kevin Stachowiak for his comments that have helped clarify this work
and to Matt for his limitless patience and support throughout this endeavor. Finally, I
wish to thank the CITY Project Team and the young men who took the time to participate

in this study.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. vi
CHAPTER 1

INTRODUCTION ............................................................................................................... l

Epidemiology of HIV/AIDS in the United States ............................................................ 1

Prevention among HIV-infected People ........................................................................... 2

HIV-risk Behaviors among HIV-infected Men ................................................................ 6

Risk & Protective Factors .............................................................................................. 1 1

Substance use during sex ...................................................................................... ll

Trading Sex .......................................................................................................... 12

Social Support ...................................................................................................... 14

Peer Norms ........................................................................................................... 15

Education .............................................................................................................. 16

Employment ......................................................................................................... 17

Exposure to Prevention Messages ........................................................................ 18

Summary ........................................................................................................................ 18

Research Questions & Hypotheses ................................................................................ 19
CHAPTER 2

METHODS ........................................................................................................................ 22

Sample ............................................................................................................................. 22

Procedure ........................................................................................................................ 23

Measures ......................................................................................................................... 24

HIV serostatus ...................................................................................................... 24

Age ........................................................................................................................ 25

Race and ethnic background ................................................................................. 25

Educational attainment .......................................................................................... 25

Employment .......................................................................................................... 25

Trading Sex ........................................................................................................... 26

UAI with male partners ......................................................................................... 26

Unprotected sex with female partners ................................................................... 26

Total number of sex partners ................................................................................ 27

UAI with Main and Non-main Partners ................................................................ 27

Sex under the inﬂuence of Alcohol or Drugs with Partners ................................. 28

Knowledge of Partners’ HIV Status ..................................................................... 28

Communication about Safer Sex .......................................................................... 28

Peer Norms about Condoms ................................................................................. 28

Social Support for Safer Sex ................................................................................. 29

Exposure to HIV Prevention Interventions ........................................................... 29

Data Analysis Plan .......................................................................................................... 30

iv

CHAPTER 3

RESULTS ....................................................................................................................... 32
CHAPTER 4
DISCUSSION ................................................................................................................. 37
Similarities among HIV-positive and HIV-negative MSM .................................. 37
Differences between HIV-positive and HIV-negative MSM ............................... 40
Limitations ............................................................................................................ 45
Implications for Prevention ................................................................................... 47
CONCLUSIONS ............................................................................................................. 49
APPENDICES ................................................................................................................ 50
REFERENCES ............................................................................................................... 76
FOOTNOTES ................................................................................................................. 84

LIST OF TABLES

Table 1 Cities and Racial Strata (N=83 75) ....................................................................... 69
Table 2 Descriptive Statistics of Study Variables among HIV-negative and HIV-positive
young MSM ........................................................................................................................ 70
Table 3 High Risk Sexual Behaviors ofHIV-negative young MSM versus those who are
HIV-positive ....................................................................................................................... 71
Table 4 Bivariate Correlations Matrix for Main Study Variables ..................................... 73

Table 5 Sequential Logistic Regression Analysis Predicting Unprotected Anal
Intercourse among HIV-positive young MSM (N = 1 48) .................................................... 74

Table 6 Sequential Logistic Regression Analysis Predicting Unprotected Anal
Intercourse among H] V -negative young MSM UV: 72 79) .................................... .75

vi

Chapter 1
Introduction

Acquired Immunodeﬁciency Syndrome (AIDS) was ﬁrst recognized in 1981.
AIDS is a fatal disease that is associated with a severely poor immune system
(immunosupression) and one or more opportunistic infections. AIDS is caused by the
Human Immunodeﬁciency Virus (HIV). Researchers are uncertain as to how the virus
entered the human population, but the ﬁrst cases in the United States were among men
who have sex with men (MSM) (Stine, 2002). On average, the HIV virus does not
progress to AIDS for 10 years after seroconversion (CDC, 1998).

According to the Centers for Disease Control (CDC, 1999), HIV is most
commonly transmitted through blood and semen. Transmission of the virus occurs when
HIV-infected blood or semen enters the bloodstream of someone who is uninfected. The
majority of people contract the virus through unprotected vaginal or anal sex, or sharing
syringes during drug use (CDC, 2007). Unprotected anal sex, in particular, is the most
dangerous sexual activity for contracting HIV since the walls of the rectum are very thin,
which allow the virus to enter the bloodstream during sexual intercourse (Stine, 2002).
Thus, men who have sex with men (MSM) are at high risk for HIV infection.

Epidemiology of HI V/AIDS in the United States

At the end of 2005, it has been estimated that 1,205,969 individuals were living
with HIV or AIDS in the United States (CDC, 2007). According to the most recent CDC
surveillance report (2007), 55,000-60,000 people are becoming infected with HIV/AIDS
every year. In terms of incidence rates, disparities in gender, age, and ethnicity exist. For

instance, 71% of new HIV diagnoses made between 2001-2005 were among men. The

highest rate of new HIV diagnoses were among men aged 25-44 (62%). Although young
men under 25 years old accounted for only 14% of new cases, the rate of new HIV
diagnoses among this age group has steadily increased since 2001. African Americans
experienced the highest rate of new HIV diagnoses (51%), followed by whites (29%) and
ﬁnally Hispanics (18%). Regarding racial trends, rates of new HIV cases among African
Americans and Hispanics have substantially increased, while infections among whites
have decreased over the past 5 years.

By 2005, over 550,000 people had died of AIDS-related causes in the United
States (CDC, 2007). AIDS related mortality reached its peak in 1994 but sharply dropped
between 1996 and 2000 (Karon, Fleming, Steketee, & De Cook, 2001). Recent data
suggests that this downward trend reached a plateau in 2001 and has remained stable
through 2005 (CDC, 2005a). This decrease in mortality is most likely attributable to
advances in antiviral medication (CDC, 2005b).

In summary, the prevalence of HIV and AIDS is disproportionately elevated
among MSM and racial minorities. Although the rates of new HIV infections seem to
have stabilized over the past decade, the proportion of these cases are growing among
younger MSM. With thousands of Americans becoming infected every year, HIV/AIDS
remains an important public health issue. Furthermore, the number of AIDS-related
deaths is substantially lower than the number of new infections. Thus, there are more
people than ever with this disease who are living in the US. population.

Prevention among HIV- infected People
Prevention initiatives have traditionally focused on reducing HIV-risk among

uninfected individuals (Janssen et al., 2001; Kalichman, Kelly, & Rompa, 1997).

Prevention professionals have typically assumed that HIV testing and a one-Shot post-test
counseling session was sufﬁcient to signiﬁcantly reduce the sexual risk behaviors of men
who are diagnosed with HIV or AIDS (Norton, Miller, & Johnson, 1997). Unfortunately,
several obstacles have stood in the way Of HIV testing as a prevention initiative for
people living with HIV/AIDS. First, when the HIV test became widely available in 1985
(Wolitski et al., 2006) there was no cure or treatment for HIV, which left little incentive
for people to get tested. Second, advocacy organizations for gays and lesbians, and
people living with HIV/AIDS such as the ACLU and the Gay Men’s Health Crisis did not
readily embrace HIV testing because of its initial mis-uses (Schiltz & Sandfort, 2000).
For example, health insurance companies made attempts to screen applicants for HIV in
order to avoid paying for the medical treatment of these individuals (Malinowsky &
Perry, 1989). Life insurance companies followed suit, requiring individuals to get tested
for HIV in order to obtain coverage (Malinowsky & Perry, 1989). Finally, the general
public held many misconceptions concerning the spread of HIV. Public opinion polls in
the mid-1980’s revealed that the majority of Americans believed the virus could be
transmitted through non-sexual, casual contact (Herek & Glunt, 1988). In turn, people
living with HIV/AIDS were feared and stigmatized. Thus, individuals had nothing to gain
but a lot to lose by obtaining an HIV test.

However, in 1987, Azidothymidine (AZT), the ﬁrst treatment for HIV was
approved by the Food and Drug Administration (FDA), and other effective treatments
that delayed the progression of HIV to AIDS became available thereafter (Beck-Sague,
2004). Voluntary HIV testing began to increase because these new treatments provided

an incentive for individuals to get tested. Although the testing rates have increased over

the past 2 decades, the preventative utility of HIV testing and counseling with individuals
diagnosed with HIV/AIDS remains in question. Most importantly, as discussed later, the

data remain equivocal regarding whether or not people change as a result of learning that
they are HIV-positive.

Advances in treatment for AIDS also have implications for prevention. Since the
mid 1990’s, the life expectancy of HIV-infected individuals has vastly improved due to
new and improved combination therapies (Beck-Sague, 2004). With HIV—positive people
leading longer and healthier lives, they can also engage in sexual activity for years
following HIV infection. Unfortunately, when HIV-positive individuals have sexual
contact with serodisconcordant partners, they place individuals at risk for contracting the
virus if condoms are not used consistently and properly. Since it has been estimated that
there are now over 1 million people in the United States living with HIV (Glynn &
Rhodes, 2005), if these individuals are engaging in risky sexual behaviors, it is an
important public health priority to initiate prevention activities with this population in
order to control and eradicate the epidemic.

In 2002, in response to the need to include HIV-positive individuals in risk
reduction initiatives, the CDC unveiled the Serostatus Approach to Fighting the HIV
Epidemic (SAFE) (CDC, 2003). This plan included goals to increase the number of HIV-
infected persons who know their serostatus, increase the use of health care and prevention
services, and increase high—quality care and treatment. However, for the ﬁrst time in
history, the CDC incorporated two additional goals speciﬁc to HIV-infected individuals:
1) increase adherence to antiviral therapy among individuals with HIV and 2) increase

the number of individuals with HIV who adopt and sustain HIV and STD risk reduction

behavior. In 2003, the CDC continued to emphasize the need to target HIV-infected
individuals for prevention in their Advancing HIV Prevention (AHP) plan (CDC, 2003).
AHP included plans to initiate prevention programs for HIV-positive individuals in
public health departments and AIDS service organizations throughout the country.

Although prevention programs for HIV-positive individuals are now being
disseminated across the nation, these programs are not targeted to all of the various
subgroups among the HIV-positive population. In particular, few innovations have been
tailored to meet the speciﬁc needs of young MSM or young MSM of color. For example.
in 2006, the CDC announced 18 “Best-Evidence Interventions” (CDC, 2006) as the most
effective HIV prevention programming available. Among these interventions, three were
designed for “high-risk” HIV-negative youth, one was for HIV-positive adult MSM, two
targeted HIV-negative MSM and the remainder was focused on heterosexual adults. Only
one program targeted HIV-positive youth. This individual-level intervention entitled
“Choosing Life: Empowerment, Action, Results” (CLEAR) is speciﬁcally targeted to
young HIV-infected men and women who have a history of substance abuse. The lack of
prevention programs for HIV-positive young MSM might be partly attributed to the fact
that we know very little about the sexual behaviors of this subgroup or the risk and
protective factors surrounding their sexual behaviors. Without such information, it may
be difﬁcult to design effective prevention initiatives to meet the prevention needs of
young HIV-infected MSM.

In summary, prevention programming has typically focused on reducing sexual
risk behaviors among HIV-negative individuals. Although at one time it was assumed

that testing, diagnosis, and post-test counseling was sufﬁcient as prevention for HIV-

infected individuals, data Show this as insufﬁcient. Increased prevention efforts are
needed in order to reduce sexual risk with this population. Such interventions are
beginning to be developed and disseminated to adult HIV-positive men. However, little
has been done to address the prevention needs of younger MSM living with HIV/AIDS,
especially among ethnic and racial minorities. Research exploring the sexual risk patterns
of young HIV-infected MSM may facilitate the development of such prevention
initiatives.
HIV-risk Behaviors among HI V -infected Men

Risky sexual behaviors seem to vary among age groups of HIV-positive MSM.
Generally, current research supports that adult MSM reduce their risky sexual behaviors
immediately following their HIV-positive diagnosis. For instance, in a meta-analytic
review of 27 studies that investigated the effects of HIV testing and counseling on HIV-
risk behavior, researchers found that individuals who tested HIV-positive were more
likely to decrease engagement in unprotected sexual intercourse and increase condom use
following testing and counseling when compared to HIV-negative and untested groups
(Weinhardt, Carey, Johnson, & Bickham, 1999). It should be noted that only seven of the
studies included in this meta-analysis were conducted with MSM and none exclusively
included HIV-positive young MSM. Thus, it is unclear whether these results would apply
to young MSM. Furthermore, these studies do not paint a clear picture of how long the
trends in decreased risk behaviors endure after diagnosis. The studies included in this
review only tracked HIV-positive participants for an average of 261 days. Thus, it may be
true that adult MSM decrease their risky sexual behaviors, but the duration of this trend

remains in question.

Recent research suggests that rates of unprotected sex increase over time among
HIV-positive men. For example, in a recent study that was not included in Weinhardt’s
meta-analysis, DiFranceisco, Pinkerton, Dyatlov, & Swain (2005) found. that individuals
who were diagnosed within 3 months of the study were more likely to report condom use
when compared with individuals who had been diagnosed at least 1 year prior to the
study. Another study that tracked HIV-positive adults longitudinally found that that the
rate of unprotected insertive anal intercourse with a negative or unknown status partner
sharply dropped immediately after diagnosis, but increased signiﬁcantly 12 months later
(Colfax et al., 2002). Thus, it appears that HIV-positive men decrease HIV-risk behavior
following diagnosis but these changes are not sustained over time.

Cross-sectional studies providing HIV testing to research participants have also
contributed to the understanding of sexual risk behaviors among HIV seropositive adult
MSM. The majority of these studies compare sexual risk among HIV-positive individuals
to similar samples composed of individuals who test negative. Such research has found
that HIV-infected adult MSM who are aware of their HIV seropositivity engage in
similar rates of unprotected sex as their HIV-negative peers (Deren, Beardsley, Coyle, &
Singer, 1998; George, Green, & Murphy, 1998; Kral et al., 2005; Myers, Javanbakht,
Martinez, & Obediah, 2003; Ostrow, McKiman, Klein, & DiFranceisco, 1999). There is
no compelling evidence that HIV-infected adult men engage in fewer or greater sexual
risk behaviors when compared to samples of HIV—negative adult men. In sum, adult
MSM living with HIV/AIDS seem to reduce their risky sexual behaviors following
diagnosis. However, there is accumulating evidence that engagement in sexual risk

behaviors begin returning to pre-diagnosis levels by about 12 months post-diagnosis.

Although adult MSM living with HIV/AIDS seem to decrease their sexual risk
behavior following diagnosis, it is unclear whether these behavior changes occur among
young and adolescent HIV-positive MSM. No longitudinal studies have examined the
trajectory of sexual risk behaviors after HIV diagnosis among young MSM. However, a
small number of cross-sectional studies with young HIV-infected MSM have explored
the sexual risk behaviors of these individuals. For instance, the Young Men’s Survey
(YMS) (Valleroy et al., 2000), the largest of these studies, assessed HIV-risk among
young and adolescent MSM. This multi-site study was conducted between 1994-1998
and recruited 3,492 MSM between the ages of 15-22. Of the 3,492 men sampled, 7.2%
tested HIV-positive. Only 18% of these men knew their HIV-positive status prior to
contact with the researchers. Findings from this study indicated that the young men who
were not aware of their HIV seropositivity were more likely to participate in unprotected
receptive or insertive anal intercourse during the preceding 6 months when compared
with HIV-negative men. Furthermore, the men who were aware that they were HIV-
infected were equally as likely to engage in unprotected sex as their HIV-negative peers.
Thus, this research suggests that HIV—positive young MSM continue to engage in sexual
risk behaviors regardless of their HIV serostatus knowledge. In addition, these results
contrast with ﬁndings from similar studies of adult HIV-positive MSM. While adult
MSM seem to decrease engagement in risky sexual behaviors following diagnosis, young
men who are aware they are HIV seropositive seem not to alter their risky sexual
behaviors.

High rates of risky sexual behaviors among HIV-positive young MSM were also

found in the San Francisco Young Men’s Health Study (Hays et al., 1997). This project

used a probability sample of 408 males aged 18-29 from the San Francisco Area. Of
those men who were previously tested for HIV antibodies, 17% reported they were
infected with the HIV virus. After these men were tested by the researchers, an additional
4% were found to be HIV-positive. Among men who had never been previously tested
for HIV, 10% were found to be HIV-positive by the researchers. Thus, a total of 18.7% of
the entire sample was HIV—positive with 25% being unaware of their seropositivity prior
to contact with the researchers. Sexual risk behaviors were compared between three
categories of men: 1) men who perceived themselves to be HIV-negative, 2) men who
perceived themselves to be HIV-positive, and 3) men who were not tested prior to contact
with the research staff and were unaware of their serostatus. It was found that the HIV-
positive group of men engaged in the highest rates of risky sexual behavior. For instance,
59% of those who perceived themselves to be HIV-positive reported having unprotected
receptive and/or insertive anal intercourse within the previous year. These same
behaviors were only endorsed by 35% of the men who perceived themselves to be HIV-
negative and 28% among the untested group. These ﬁndings further support the notion
that young HIV-positive MSM are engaging in signiﬁcantly elevated rates of unprotected
sex when compared to their HIV-negative peers. The rates of unprotected sex are also
higher than typically found in studies of adult men.

Studies relying on convenience samples of young MSM provide additional
evidence for high rates of risky sexual behavior among HIV-infected young men. For
instance, Hein, Dell, Futtennan, Rotheram-Borus, & Shaffer (1995) examined the sexual
risks behaviors of 793 adolescents aged 13-21 who were receiving inpatient or outpatient

medical and/or psychiatric services. Following HIV antibody testing, 9% of the sample

was determined to be HIV-positive. In comparing the participants by their serostatus, the
researchers found that HIV-positive men were signiﬁcantly more likely to engage in
insertive and receptive oral and/or anal intercourse. HIV-positive adolescents also
reported to have experienced a greater number of sexually transmitted diseases, and were
more likely to have engaged in survival sex, sex with casual partners, and use condoms
infrequently when compared to HIV-negative men. Another study that explored the
patterns of risky sexual behaviors of young HIV-positive MSM was conducted by Stein,
Rotheram-Borus, & Milbum (2005). This study employed a convenience sample of 248
adolescent MSM who were receiving services at clinical care sites in four AIDS
epicenters across the United States. Findings from this study indicated that only 17% of
248 HIV-positive young males abstained from sex. Furthermore, 23% of the participants
reported that they had never used condoms during sex within the prior 3 months. Also,
the men reported an average of seven sexual partners within the preceding 3 months.
Although this study did not utilize a comparison group of HIV-negative MSM, the rates
of sexual risk behavior reported by these men are alarmingly high.

As with most community-based research, ﬁndings from the aforementioned
studies dealing with HIV-positive youth come with signiﬁcant limitations. One major
problem is that HIV-infected young men are a hard to reach population, which makes
obtaining representative samples an enormous challenge. Thus, some researchers have
relied upon convenience samples of participants receiving medical and/or mental health
services (Hein, Dell, Futterman, Rotheram-Borus, & Shaffer, 1995; Stein, Rotheram-

Borus, Swendeman, & Milbum, 2005). However, the results of studies using a

10

community-based sample do not differ from those that employed probability samples
(e.g. Valleroy et al., 2000). Therefore, we may be more conﬁdent in these ﬁndings.

The preceding studies indicate that HIV-positive young men engage in an
alarmingly high rate of unprotected sex. In contrast with studies of adults, no research
with young HIV-infected young MSM has provided evidence that men who are aware of
their HIV-positive status reduce their high risk sexual behaviors after diagnosis. These
studies provide some evidence that young HIV—infected MSM who are aware of their
HIV status, in comparison to their HIV-negative peers, report either equal (Valleroy et
al., 2000) or higher rates of risky sexual behaviors (Hays et al., 1997; Hein et al., 1995).
Regarding men who are unaware of their HIV status, only two studies provided such
analyses and their results were inconsistent. In any case, this small body of research
provides some compelling support that young HIV-positive MSM are engaging in high
rates of unprotected sex. In turn, these young MSM are undoubtedly making a signiﬁcant
contribution to the spread of HIV/AIDS.

Risk & Protective Factors

Many of the published studies with young HIV-positive MSM have primarily
dealt with risk factors rather than protective factors for high risk sexual behaviors. These
studies have taken a deﬁcit-oriented approach to studying predictors of risky sexual
behaviors. Common risk factors that have been investigated within this population
include substance use during sex and engagement in sex trade.
Substance use during sex

Substance use presents a serious barrier to HIV prevention initiatives among

MSM due to its relationship with risky sexual behaviors. The use of these illegal

11.

substances before, or during sex is problematic because they are known to impair one’s
judgment. Thus, individuals who are “high” may fail to engage in risk reduction practices
within their sexual encounters.

Substance use among HIV-negative or unknown serostatus MSM is high. Many
MSM use drugs and alcohol during sexual activities (Stall & Purcell, 2000). Use of
alcohol, nitrite inhalants, crystal methamphetamine, cocaine, marijuana, and Viagra have
been implicated in acts of unprotected anal sex among gay and bisexual men (Hirshﬁeld,
Remien, Humberstone, Walavalkar, & Chiasson, 2004). The relationship between
substance use and high risk sexual behavior has also been reported among HIV-positive
young MSM. For instance, the Young Men’s Survey found that HIV-infected young
MSM were more likely to have sex while under the inﬂuence of crack cocaine, nitrite
inhalants, or injection drugs when compared with HIV-negative MSM (Harawa et al.,
2004). In another study of inpatient and outpatient young men receiving services at a
medical center in New York City, comparisons of HIV-positive and HIV—negative
adolescents showed that those who were HIV-positive were more likely to have engaged
in sexual activities while drunk or high (Hein et al., 1995). These studies Show that drug
and alcohol use during sex is a risk factor for unprotected sex among both HIV-negative
and positive men. Moreover, it appears that this risk may be elevated among HIV
seropositive young MSM.
Trading Sex

Trading sex (sex in exchange for money, food, shelter, drugs, or other items) is a
well documented risk factor for HIV-risk behavior among young MSM. Studies

consistently report that engagement in sex trade is associated with unprotected sex

12

(Bailey, Camlin, & Ennett, 1998; Greene, Ennett, & Ringwalt, 1999; Halcaon & Lifson,
2004; Kral, Molnar, Booth, & Watters, 1997; Whitbeck, Chen, Hoyt, Tyler, & Johnson,
2004). However, less is known about trading sex and HIV-risk among HIV-positive
young MSM. Some evidence suggests that HIV-positive youth are Signiﬁcantly more
likely to engage in trading sex when compared to those who are HIV-negative. For
instance, in a study of 8,000 adolescents across the United States, those who participated
in sex trade were almost three times as likely as those without this risk factor to be HIV-
positive (Huba et al., 2000). In the aforementioned study by Hein et al. (1995), HIV-
positive adolescents were signiﬁcantly more likely to report having participated in
trading sex when compared with HIV-negatives (32% vs. 4%). Pfeifer & Oliver (1997)
also found high rates of trading sex among HIV-infected youth. In a sample of 96
homeless youth aged 14-24 in Hollywood, California, researchers found that those who
tested HIV-positive were signiﬁcantly more likely to trade sex for money or drugs, and
also to engage in unprotected vaginal and receptive anal sex when compared to
uninfected youth.

Based on the few studies exploring trading sex among HIV-positive young MSM,
it is clear that these men are engaging in high rates of sex trade when compared to their
uninfected peers. It appears that young HIV seropositive MSM are also having
unprotected sex. However, it is not known whether HIV-positive young MSM who
engage in trading sex are more or less likely to participate in unprotected sex. Due to the
fact that most HIV-negative or unknown serostatus MSM participating in trading sex

often report high numbers of sexual partners (Haley, Roy, Leclerc, Boudreau, & Boivin,

I3

2004). HIV-infected MSM who trade sex could be contributing signiﬁcantly to the spread
of HIV if they are not practicing safer sex.
Social Support

Social support has been theorized to serve as a protective factor against negative
health behaviors. For example Cohen (1988) proposed that those who provide social
support render advice about healthy practices and risk reduction strategies, and also
promote the practice of positive health behaviors. However, researchers have failed to
ﬁnd a signiﬁcant relationship between social support and risky sexual behaviors among
HIV-negative and HIV-positive adult men. For instance, Crepaz and Marks (2002)
conducted a meta-analytic study on various predictors of unprotected sex among HIV-
positive individuals. The researchers concluded that, overall, social support was not a
signiﬁcant predictor of risky sexual behavior. Studies speciﬁcally among young HIV-
infected MSM have been less conclusive in regards to the role of social support in sexual
risk taking behaviors. For example, the San Francisco Young Men’s Health survey did
not ﬁnd a signiﬁcant relationship between social support and risky sexual behavior
among participants; social support did not predict unprotected anal intercourse for those
who perceived themselves to be HIV-positive or HIV-negative (Hays et al., 1997).
Another study comparing HIV-positive and HIV-negative youths aged 12 to 21 receiving
inpatient and outpatient medical services found that the role of social support did not
predict differences in sexual risk behavior between HIV-positive and HIV-negative
participants (Hein et al., 1995). Consistent with these ﬁndings, Stein, Rotheram-Borus.
Swendeman, & Milbum (2005) did not ﬁnd that social support predicted sexual

transmission risk behaviors in a sample of 342 youths living with HIV/AIDS in clinical

14

sites across four major cities in the United States. Conversely, Naar-King, et al. (2006)
examined the relationship between social support for safer sex and unprotected condom
use among 60 HIV-positive youth receiving medical services. The researchers found that
participants who had reported greater support for safer sex were more likely to use
condoms. However, general social support was not found to be predictive of safer sexual
behaviors.

Although few studies found evidence for social support as a predictor of high risk
sexual behavior, this issue merits further exploration because these researchers have
deﬁned social support differently. The studies above provide evidence that emotional
support (Hein et al., 1995) and general support (Stein et al., 2005), did not predict sexual
risk among young HIV-infected men. However, social support speciﬁc to sexual
behaviors has been shown to signiﬁcantly predict condom use (Naar-King et al., 2006).
Peer Norms

Peer norms have also been found to be protective factors against negative
outcomes for HIV-positive youth. There is an abundance of research suggesting that peer
norms play a protective role against risky sexual behaviors among both adult and young
MSM (Kalichman, Kelly, & St. Lawrence, 1990; Somlai, Kalichman, & Bagnall, 2001).
Fisher (1988) proposed that individuals tend to practice safe sex to the degree that it
conforms to the norms of their particular reference group. However, the role of peer
norms in protecting against high risk sexual behaviors among young HIV-positive MSM
is unclear. For instance, the San Francisco Young Men’s Health Study did not ﬁnd social

norms to be predictive of unprotected anal intercourse for HIV-positive or untested young

15

men (Hays et al., 1997). Countering this research, Stein et. al, (2005) did ﬁnd a direct
signiﬁcant relationship between unprotected sex and negative peer norms.

The preceding two studies discussed are the only published research investigating
the relationship between peer norms and risky sexual behavior among young, HIV-
positive males. Both of these studies used different measures of peer norms, thereby
limiting our ability to draw any compelling conclusions. The San Francisco Young Men‘s
Health Study only used one item to measure perceived social norms regarding safe sex.
The latter study by Hayes et al. (1997) employed the Inventory of Socially Supportive
Behaviors, which was a general index of negative peer norms regarding alcohol use, drug
use, and promiscuity. Further investigation into the role of peer norms in sexual
behaviors of young HIV-positive men is needed.

Education

The relationship between educational attainment and participation in unprotected
sex has been documented in a variety of different populations (Hasnain, Levy, Mensah, &
Sinacore, 2007; Kelly, Sikkema, Winett, Solomon, & et al., 1995; Paasche-Orlow,
Clarke, Hebert, Ray, & Stein, 2005). While it is clear that education plays a role in risky
sex among HIV-negative MSM, it is unclear whether this same relationship is true for
HIV-positive MSM. In the study conducted by Hays et al. (1997) with young MSM,
HIV-negative men who received fewer years of formal education were more likely to
engage in UAI. However, an opposite pattern emerged among the HIV-positive men;
increased educational attainment was associated with higher rates of UAI. Contradicting

these ﬁndings, one study of adult HIV+ MSM reported that those without education

16

beyond high school were more likely to have participated in UAI during the previous
year (Denning & Campsmith, 2005).

Educational attainment may play a role in engagement of unprotected sex for
several reasons. First, men who are less educated may have more difﬁculty accessing and
comprehending HIV prevention information. Second, less educated people are more
likely to hold misconceptions about what sexual behaviors are safe, and which are not
(LeBlanc, 1993; Salmon, Wooten, Gentry, Cole, & Kroger, 1996). Thus, these
individuals may make themselves vulnerable to HIV infection by na'ive engagement in
risky sex. Last, less educated MSM may have poor sexual communication skills. HIV-
positive men who are able to effectively discuss issues related to safer sex with their
partners have been found to be less likely to engage in unprotected intercourse (Semple.
Patterson, & Grant, 2000). More research is needed in order to better understand the role
of education in rates of high risk sexual behaviors among young HIV-positive MSM.
Employment

It is well documented that MSM who are HIV-infected are less likely to be
employed when compared to HIV-negative MSM (Harawa et al., 2004). This is
especially the case with HIV-positive men who are at more advanced stages of HIV-
infection and are physically incapable of working (Kass, Munoz, Chen, Zucconi, & Bing.
1994; Vitry-Henry, Penalba, Beguinot, & Deschamps, 2002). Less is known, however,
about the relationship between employment and risky sexual behaviors among young
HIV-infected MSM. In fact, only one study has explored the relationship of unprotected
sex and unemployment among HIV-positive MSM (Parsons, Halkitis, Wolitski, Gomez.

& Seropositive Urban Men's Study, 2003). However, the authors of this study reported a

17

null relationship between employment status and risky sexual behaviors. Currently, there
is no published research exploring the role of employment status and sexual behaviors
among younger samples of MSM. More research is critically needed in order to better
understand employment status and how it could impact HIV-positive men’s sexual
behaviors.

Exposure to H] V pre vention Messages.

Exposure to HIV prevention messages may serve as a protective factor against
HIV-risk behaviors among MSM. Theory-based prevention initiatives have been
successful at reaching gay and bisexual men (Goldbaum et al., 1998) and have been
implicated in reducing risky sexual behaviors (Herbst et al., 2005). However, it is not
known whether exposure to HIV prevention messages has the same efﬁcacy in reducing
unprotected sex with young HIV-positive MSM.

Summary

Young MSM are becoming infected with HIV at a disproportionately higher rate
than other groups within the US population. Young HIV-positive MSM, in particular.
may be at high risk for transmitting the virus to others. Yet, prevention programmers and
researchers have typically focused on the sexual risk behaviors of HIV-negative
individuals. Historically, the primary vehicle for prevention with HIV-positive people has
been through HIV testing and one-time pre-test/post-test counseling. Unfortunately,
research suggests that testing and counseling alone is insufﬁcient to have a long-tenn
impact in reducing risky sexual behaviors.

In terms of research on young HIV-infected MSM, only a few studies have

explored the patterns of sexual risk behaviors of this population. The few studies that

18

have been conducted with this population have produced contradictory results.
Furthermore, little is known about the role of risk and protective factors for unprotected
sex among young HIV-positive MSM. With thousands of new cases of HIV infection
each year, it Should be an important public health priority to develop prevention
programming for this population. More research is needed to better understand the risk
proﬁles of young HIV-positive MSM. Such information may be useful in creating
prevention programs that are speciﬁcally tailored to the needs of these men.
Research Questions & Hypotheses

In an attempt to increase our understanding of HIV-positive young MSM, the
proposed study will: 1) describe the sexual risk behaviors of HIV-infected young MSM,
2) compare rates of engagement in HIV-risk behaviors between HIV-negative and HIV-
positive young MSM, 3) identify risk and protective factors associated with unprotected
sex among HIV-positive young MSM, and 4) explore the degree to which risk and
protective factors for unsafe sex among HIV-positive young MSM explain unsafe sex
among HIV-negative young men.

The proposed study will address the following hypotheses and exploratory
research questions:

Q1: DO HIV-positive and negative young MSM engage in Similar risk behaviors

and at similar rates?

HI: Current research supports the hypotheses that young HIV-infected MSM will

engage in equal or higher rates Of unprotected anal intercourse when compared to

HIV-negative MSM. Consistent with the available literature, I hypothesize that

19

rates of high risk sexual behavior will not differentiate HIV-positive and. HIV-
negative participants.

Q2: What factors encourage and protect against engagement in unprotected anal
intercourse among young HIV-positive men?

H2: Among HIV-positive young men, as educational attainment increases, the
proportion of men who report unprotected anal intercourse will decrease,
controlling for age.

H3: Self-identiﬁed HIV-positive young MSM reporting employment will be less
likely to engage in risky sexual behaviors than their HIV-positive peers who are
unemployed.

H4: Research suggests that individuals tend to practice safer sex to the degree that
it conforms to the norms of their reference group (Fisher, 1988). Therefore, young
MSM who identify as HIV-positive and who have greater peer support for safer
sex will be less likely to participate in unprotected anal sex than young HIV-
positive MSM who have lower scores on peer norms.

H5: Social support has been theorized to serve as a protective factor against
negative health behaviors (Cohen, 1988). Thus, young MSM who identify as
HIV-positive and who have greater social support for safer sex will be less likely
to report unprotected anal intercourse than HIV-infected young MSM who report
lower rates of social support.

H6: HIV-positive young MSM who used alcohol or drugs during their last sexual
encounter will be more likely to engage in unprotected sex than HIV-positive

young MSM who did not use substances during their last sexual contact.

20

H7: Young MSM who are HIV-positive and who have participated in trading sex
within the past year will be more likely to engage in unprotected anal intercourse
than HIV-positive young MSM who do not report having traded sex.

Q3: Do the same factors protect and put at risk H IV-positive and HIV-negative
young men?

There is insufﬁcient research comparing HIV-positive and HIV-negative young
MSM in terms of risk and protective factors for risky sexual behavior to formulate
hypotheses. For the purposes of this study, comparisons of risk and protection will

be pursued in an exploratory manner.

21

Chapter 2
Method

The proposed study uses data from the Community Intervention Trial for Youth
(CITY) Project. The CITY project was a 7-year randomized multi-site control trial of
interventions to reduce HIV-risk behaviors among young MSM between the ages of 15-
25. Using venue-based time-Space sampling (TSS) (Muhib et al., 2001) CITY recruited a
community based, multi-ethnic probability sample of young MSM from 13 different
urban locales across the United States. Data were collected on young men’s sexual risk
behaviors and predictors of risk via an anonymous interview.

Sample

To be eligible to participate in the CITY Project, men had to be between the ages
of 15 to 25 years and report that they had engaged in sexual activity with another man in
the previous year. In addition, men had to meet racial eligibility criteria applicable to the
particular study site at which they were recruited. The 13 sites for the CITY Project were
geographically diverse, representative of most regions throughout the country, and had
venues where it was possible to Obtain a multi-ethnic sample of young MSM. Cities were
divided into four strata based on race/ethnicity (see Table 1).

The study utilized a venue-based application of TSS (Muhib et al., 2001). CITY
researchers at each site identiﬁed venues that were likely to be frequented by young
MSM through key informants and gay-oriented magazines and newspapers. Through a
systematic process, sites identiﬁed venue-day—time units (VDT) that were optimal for
collecting data from young men. These VDT units comprised the sampling frame. VDTs

were then randomly selected for data collection (See Muhib et. al., 2001 for a thorough

22

discussion of TSS). Randomization occurred once each month. Data were collected in
each Site from May through August in the years 1999, 2000, 2001, and 2002'.
Procedures

Data collection procedures and processes for obtaining informed consent were
reviewed and approved by the relevant human subjects review committees at each of the
13 study sites and also the Centers for Disease Control and Prevention. During data
collection events, staff systematically approached young men who exhibited physical
characteristics consistent with the race and age requirements at each site. In order to
determine eligibility, a brief screening instrument (881) was administered to each
individual who was approached. Interviewers read the questions aloud to each
respondent. Participants were eligible to be interviewed if they were between the ages of
15 and 25, had sexual contact with a male in the previous year, were a resident of the
catchment area, and met the racial/ethnic requirements of the site. Sexual contact was
deﬁned as participation in oral sex, anal sex, or any other physical contact leading to
orgasm. Once eligibility was determined, potential respondents were asked to participate
in a 20-minute interview and informed consent was obtained. Interviewees were paid
$10-$20 (depending on the site) as an incentive for their participation in the study.
Participation was anonymous. Altogether, 37,004 men were approached and 27,361
(74%) agreed to be screened. Of those, 12,525 (46%) met the criteria for inclusion and
11,779 (94.3%) of those men agreed to be interviewed. The current study will only use
data from MSM who reported to have been tested for HIV and to have received their

results. Thus, the ﬁnal analytic sample contains 8,384 young MSM.

23

Measures

The anonymous face-to-face interview protocol appears in Appendix A. Only
those measures that were used in the current analyses are described.
HIV Serostat us

HIV status was assessed by ﬁrst asking “Have you been tested for HIV”? If the
respondent replied afﬁnnatively, he was asked, “Would you be willing to tell us the
results of the last test”? Responses were coded into 7 categories: 1 (HIV-positive), 2
(HIV-negative), 3 (never tested/unknown), 4 (indeterminate), 5 (didn’t get results), 6
(don’t know), and 7 (refused).

For the current analysis, HIV status was dichotomized into an HIV-negative
group and an HIV-positive group. A total of 8,932 men reported a negative HIV-test
result and 163 were HIV-positive. Those who were not tested (n = 2274), who didn’t
return to get their results after testing (n = 180), who didn’t know their serostatus (n =
103), or who refused (n = 51) were excluded from all analyses. Fifteen men reported to
have received an indeterminate test result and were added to the HIV-positive sample.
This was done because the men with indeterminate HIV status were similar to HIV-
positive men after examining their rates of risky sexual behaviors. Also, under the CDC
guidelines for HIV testing and counseling, individuals who receive an indeterminate test
result and report a history of high risk sexual activity should receive post-test counseling
similar to those who receive a seropositive test result (CDC, 2001) because these men
report risky sexual behaviors3. After adding these participants, the sample of HIV-
positive men included 178 men.

Soda-demographics

24

Age; Young men were asked for their date of birth and also how many years old
they were. For the purposes of all analyses, age was treated as a continuous variable
measured in years. The mean age of the sample was 22 (sd = 2.26) and ranged from 15
to 25.

Race and ethnic background. Race/ethnicity was measured by two questions:
“What is your ethnicity?” and “How would you describe your racial background?” The
ﬁrst question was coded as 1 (Hispanic or Latino) or 2 (Not Hispanic or Latino). Racial
background was coded as Black, Latino, Asian American/Paciﬁc Islander, White,
American Indian, Mixed, or unknown. In preparing these data for analysis, Asian
American/Paciﬁc Islander, American Indian, multi-racial, and men reporting their race to
be unknown were combined into a single variable named “other race.” Dummy variables
were then constructed with “other race” serving as the reference category and separate
variables were created for Black, White, and Latino.

Educational attainment. Young men were asked to report how many years of
school they had completed. For analyses, education was coded into 5 categories: 0-6
years, some high school (7-11 years), high school graduate ( 12 years), some college (13-
15 years), and college graduate (16 or more years).

Employment. Employment was assessed and coded into the following categories:
Full time, part time, not working because of disability, not employed but looking for
work, not employed and not looking for work, or not employed but receiving
unemployment beneﬁts. For the current study, this variable was dichotomized so that a
code of 0 reﬂected not being employed and a code of 1 indicated part or full-time

employment.

25

HIV-risk Behaviors

HIV risk sexual behaviors were assessed for three different time frames. MSM
were asked about their sexual risk practices within the last year, within the past three
months, and during their last sexual encounter.

Sexual risk behaviors within the past year.

Trading sex. Participants were asked if they had traded sex for money, drugs, or
anything valuable (i.e. food, or a place to stay). Responses were coded 1 for yes, and 0
for no.

Sexual risk behaviors in the past three months.

UAI with male partners. This measure assessed whether the respondent had
engaged in UAI within the previous three months. If the respondent endorsed
participating in anal sex, he was asked to report the number of occasions in which he was
the receptive partner and the number of times when he was the insertive partner. The
men were also asked to report the number of times that condoms were not used during
these sexual encounters. An example of items from this measure include “In the past
three months, how many times have you had anal sex with men, where you were the
inserting partner, and you grin—m use a condom”? Those respondents who engaged in
anal sex but did not use a condom were coded 1 on this variable. Men reporting one or
more occasions of receptive or insertive anal sex without a condom were coded 0. Those
who did not report having participated in any anal sex within the previous three months
were excluded from analysis.

Unprotected sex with female partners. Respondents reported whether they had

engaged in vaginal sex or anal sex with women within the past three months. If so,

26

participants were also asked whether condoms were used “from start to ﬁnish.” For the
current analyses, a variable was created to reﬂect whether the participant had or had not
engaged in any unprotected sex with a woman. A code of 1 indicates that a man had anal
or vaginal sex without a condom. A code of 0 reﬂects no unprotected sex with a woman.
Those who had not had any sexual contact with women were excluded from analysis.

Total number of sex partners. Men were asked to report the number of different
male partners with which they had any anal sex without a condom within the past three
months. In order to reduce bias due to extreme outliers, this variable was truncated into
three categories: 0 partners, 1 partner, or more than two partners.

Most recent sexual encounter.

UAI with Main and Non-main Partners. Men were asked a series of questions
regarding their sexual activities with main and non-main partners during their last sexual
encounter. For this study, a main partner was deﬁned as, “any man you consider to be
your main male partner, like a lover or a boyfriend” while non-main partners were
described as “any man you do not consider to be your main partner.” For each partner
type, men were then asked whether they had engaged in either receptive or insertive anal
sex within the past year. Responses were coded 1 for yes, and 0 for no. If yes, they were
asked whether they (or their partners) used a condom “from start to ﬁnish” during the last
time they had sex. Responses were coded 1 for yes, and 0 for no.

For the current analyses, two variables were created: “UAI with Main Partners”
and “UAI with Non-main Partners.” For “UAI with Main Partners,” a respondent was
assigned a 1 if he reported less than perfect condom use with main partners and a 0 was

given to those using condoms 100% of the time. Men who did not have any main partners

27

were excluded from analysis. For “UAI with Non-main Partners,” men reporting
inconsistent condom use were coded as 1 while those who used condoms during every
sexual encounter were coded as 0. Those who did not have a non-main partner were
excluded from analysis.

Sex under the Inﬂuence of Alcohol or Drugs with Partners. This dichotomous
variable assessed engagement in sex while under the inﬂuence of alcohol and/or
recreational drugs. For both non-main and main partners, respondents were asked “Were
you high on drugs or alcohol the last time you had sex with him?” Responses were coded
0 for no and I for yes.

In order to prepare this variable for regression analyses, a new variable was
. created to reﬂect whether the participant had sex while high with main partners and/or
non-main partners. Thus, if a participant said yes to using alcohol or drugs to either of
these questions, he was assigned a 1. Conversely, if he replied no to both of these
questions, he was given a value of 0.

Knowledge of Partners’ HIV Status. The question “The last time you had sex
with him, did you have an idea what his HIV status might be?” was asked regarding main
and/or non-main partners. Reponses were coded 0 for no and l for yes.

Communication about Safer Sex. In reference to main and/or non-main partner,
men were asked “And the last, most recent time you had any kind of sexual contact with
him, did you communicate with him in any way about using a condom when you had sex
that time?” A yes response was coded as I and no was 0.

Peer Norms about Condoms

28

Four questions determined how the participants’ friends felt about using condoms.
These items were adapted from a longer scale developed by Choi, Coates, Catania, Lew,
& Chow (1995). Example items include “Most of my friends are using condoms these
days when they have anal sex” and “Most of my friends think you Should always use
condoms when having anal sex with a new partner.” Participants were asked to rate how
much they agreed or disagreed with these statements using a Likert scale ranging from 1,
agree strongly, to 4, disagree strongly. These scores were combined into a single index
ranging from 4 to 16 with higher scores reﬂecting stronger peer norms for condom use.
Only the scores of those who answered at least three of these questions were included in
the analyses. For this sample, the mean for peer norms about condoms was 12.94 (sd =
3). Cronbach’s internal consistency alpha was adequate (a =.75).
Social Support

A four question index of available social support for safer sexual behaviors was
included in the interview. These items were adopted from Miller (1995). Respondents
rated the extent of available social support regarding issues related to sex, such as having
someone with whom they could share concerns about AIDS, on a 5-point scale ranging
from 1=none of the time to 5=always or all of the time. Item responses were summed so
that scores could range from 4-20. High scores indicate high levels of sexual social
support. Only the scores of those respondents who provided answers to at least 4 of these
questions were included in the analyses. The mean score was 15.86 (Sd = 4). Cronbach’s
internal consistency for this sample was acceptable (or = .83).

Exposure to HIV Prevention Interventions

29

This measure included four items assessing the degree of exposure to AIDS
prevention. Each item described a prevention initiative. Examples of AIDS prevention
interventions were workshops, social events/parties, education or outreach, ﬂyers or
pamphlets, and radio, newspaper, or TV marketing. For each item on this measure,
participants were asked if they were exposed to the intervention. For example, one of the
items read “In the past six months, have you participated in any workshops about AIDS
prevention.” For the current analysis, a new variable ranging from 0 to 4 was created
reﬂecting the total number of intervention exposure types. For example, a respondent
who reported any exposure to workshops and social events/parties but did not endorse
exposure to the other two intervention types received a value of 2 for this variable.

Data Analysis

To prepare data for analyses, the frequency distributions for all study variables
were examined to identify potential problems (e. g. outliers and missing data). Following
examination of the data for missing and distributional properties, descriptive statistics
were run, including means and standard deviations, in order to examine the
characteristics of the sample and to compare the characteristics of the positive and
negative men on age, race, education, employment, HIV status, sexual behaviors, social
support, peer norms, substance use, and engagement in trading sex. See appendix B for a
detailed description of the data cleaning process.

In addressing research question I, the reported rate of several HIV-risk behaviors
between the positive and negative men in the sample were compared using Pearson’s
Chi-square test of independence for categorical variables and a dependent t-test for

continuous variables. For research question II, only respondents who self-identiﬁed as

30

HIV-positive were included in these analyses. Sequential logistic regression, in which the
dependent variable was UAI within the past three months was employed. In the Ist step
of this analysis, I controlled for the effects of age, race, education, and employment. The
2nd step included social norms, social support, exposure to prevention messages,
substance use, and trading sex. The Chi-square likelihood ratio determined the overall
signiﬁcance of the model. In addition, the Wald statistic identiﬁed which predictor
variables made a signiﬁcant impact on the dependent variable after controlling for age
and race. Odds ratios and conﬁdence intervals were also included.

Finally, in order to explore potential differences in risk and protective factors for
unprotected sex between HIV-positive and HIV-negative participants, an identical
logistic regression model was run, including only the HIV-negative participants. Identical
test statistics used for the previous analyses were employed to determine the signiﬁcance
of the model and the impact of each individual predictor variable on the dependent
variable. In exploring the differences between the two models, the model ﬁt of the data,

the direction of effects, and whether the models had similar strength were investigated.

31

Chapter 3
Results

Descriptive analyses of predictor and outcome variables were performed.
Comparisons were made on socio-demographic variables between the two serostatus
groups. The chi-square likelihood ratio test and t-test were used in the analysis of
categorical and continuous variables. As table 2 indicates, HIV-positive men were
signiﬁcantly older than HIV-negative men. There were also signiﬁcant race differences
between the two groups. For instance, a much larger percentage of the men in the HIV-
pOSitive sample were African American and the minority were white. There were also
differences in rates of employment: HIV-positive men were less likely to be working
when compared to those in the HIV-negative group. Finally, there were no differences
between the two groups regarding sexual orientation with most of the men in the sample
reporting to be gay or bisexual.

Chi-square tests for categorical variables and t-tests for continuous variables were
used to identify which HIV-risk behaviors were associated with HIV-status. The
likelihood ratio for chi-square was used to determine statistically signiﬁcant relationships
between the variables. To detect whether differences between means for continuous
variables were statistically signiﬁcant, the obtained t was compared to the critical value
(Keppel & Wickens, 2004). As shown in table 3, HIV-positive men reported Signiﬁcantly
greater sexual risk on three variables. HIV-positive men were more likely to have traded
sex within the past year when compared to HIV-negative MSM. In addition, having sex
while under the inﬂuence of alcohol or drugs was also more common among these men.

However, this was only true within sexual encounters with main partners: having sex

32

while intoxicated with non-main partners was not signiﬁcantly associated with UAI.
Finally, HIV-positive young men reported to have had UAI with a greater number Of
partners when compared to their HIV-negative peers. On the remaining nine variables,
there were no signiﬁcant differences in risk behaviors between the two groups. It should
be noted that the relationship between HIV-status and UAI or UVI with women could not
be analyzed because the number of HIV-positive men engaging in this activity was too
small. Tabachnick & Fidell (2007) recommend a minimum of 5 observations per cell
when performing the chi-square likelihood ratio test.

Sequential logistic regression was employed to examine the impact of risk and
protective factors on engagement in UAI among HIV-positive men. Bivariate correlations
for all variables included in the following regression models are reported in table 4.
Sequential logistic regression was chosen for these analyses because binary variables can
be used as the dependent and predictors can include both categorical and/or continuous
variables. Logistic regression is also robust to a variety of assumptions: it does not
require normal distributions, does not assume linearity of relationships between variables,
and does not presuppose homoscedasticity (Tabachnick & F idell, 2006)..

For the ﬁrst logistic regression analysis, only HIV-positive MSM were selected.
Any UAI in the past three months served as a dichotomous outcome while socio-
demographic variables were entered in the ﬁrst step. These indicators included age, race.
education, and employmentz. The second step included social norms, social support,
exposure to prevention, trading sex, and substance use during sex as predictor variables.

Among the HIV-positive sample, 11 cases with missing data on predictor

variables were deleted. An additional 29 men were excluded from analysis since they did

33

not report engaging in any anal sex within the previous three months. Thus, 131 cases
were available for analysis. Evaluation of adequacy of expected frequencies for
categorical predictors did not warrant restricting the model goodness-of-ﬁt tests. No
serious violation of linearity in the logit was observed.

The chi-square likelihood ratio test was used to test the differences between the
steps entered in the sequential logistic analyses. The model chi-square is reported for
each step, which determined whether the likelihood ratios (-2LL) Signiﬁcantly differed
between the two steps. Although there is no statistic in logistic regression that is directly
equivalent to a linear regression R2 , Nalgelkerke R2 provides a measure indicating the
strength of model ﬁt (Field, 2005). The Wald test was used in order to test the
signiﬁcance of individual model parameters.

Table 5 shows regression coefﬁcients, Wald statistics, odds ratios, and 95%
conﬁdence intervals for odds ratios for each step within the model. As seen in the table,
the ﬁrst step containing age, race, education, and employment was not statistically
signiﬁcant and therefore was not associated with UAI among the sample of HIV-positive
men. However, after adding peer norms, social support, exposure to prevention, trading
sex, and substance use, the full model was statistically signiﬁcant. Thus, the predictors, as
a set, reliably distinguished between those HIV-positive men engaging in UAI and those
who did not report UAI. According to Wald criterion, both peer norms and social support
were associated with engagement in UAI within the past three months. Young MSM
reporting greater peer norms for condom use were 1.13 times less likely to have engaged
in UAI. Also, those who reported increased social support for safer sex were 1.17 times

likely to have abstained from UAI within the past three months.

34

The next analysis sought to determine whether the same factors protected or put at
risk HIV-negative young MSM. In accord, an identical sequential logistic regression as
described above was performed with only the HIV-negative respondents. Next, the
logistic regression analysis with the HIV-positive men was compared to the analysis with
the HIV-negative men in order to determine whether different Signiﬁcant predictors
emerged between the two serostatus groups.

The logistic regression analysis with HIV-negative respondents included 6,667
men after 258 cases were removed because of missing data and 1,279 men were excluded
because they had not had anal sex within the past three months. Regression coefﬁcients,
Wald statistics, odds ratios, and 95% conﬁdence intervals for odds ratios for the full
model appear in Table 6. The ﬁrst set of socio-demographic variables, as a set,
signiﬁcantly predicted UAI. However, race and employment were the only statistically
signiﬁcant predictors of UAI. Both Hispanic and Black MSM were more likely to have
abstained from UAI than those in the “other race” category. Furthermore, employed men
were less likely to have had UAI. After the addition of the second step, the full model
with all the variables signiﬁcantly predicted UAI. Race remained signiﬁcant after the
second step was added, although the effect of employment on UAI did not. Within the
second step, peer norms were the only signiﬁcant variable associated with UAI. HIV-
negative men who reported greater peer norms for condom use were 1.12 times less
likely to have had UAI within the prior three months.

In comparing the two logistic regression models, UAI was explained by mostly
different variables. The only common factor associated with UAI between the HIV-

positive and the HIV-negative groups was peer norms for condom use. Those reporting

35

positive peer norms for condoms were less likely to engage in UAI regardless of HIV
status. Social support for safer sex was a Signiﬁcant predictor of UAI for the HIV-
positive men but this relationship was insigniﬁcant for HIV-negative men. Conversely,
race was only a signiﬁcant predictor of UAI for the HIV-negative men. Finally, in
comparing the Nalgelkerke R2 statistics between the two logistic regression analyses, it
appears that the predictor variables better explain the variance in UAI for the HIV-

positive group, then for the HIV-negative group.

36

Chapter 4
Discussion

Until recently, prevention initiatives have focused their efforts on HIV-negative
men under the guiding assumption that individuals will change their high risk sexual
behaviors in light of receiving an HIV-positive diagnosis. However, ﬁndings from this
study suggest that HIV-positive MSM are an important target for prevention
interventions. Young men living with HIV/AIDS are indeed having unsafe sex. Our
ﬁndings indicate that nearly 40% of HIV-infected men and 34% of HIV-negative MSM
reported to have engaged in UAI within the previous three months. Furthermore, we
assessed several other indicators of sexual risk between the two serostatus groups and
few differences emerged. These data demonstrate that knowledge of one’s HIV status
may not contribute to the sexual decision making among many young MSM.
Similarities among HIV-positive and HI V-negative MSM

Young men, regardless of serostatus, are developmentally similar, which may
play an important role in their sexual risk taking behaviors. One possible explanation for
the high rates of risky sex among young MSM is that they Often feel invincible to the
potential negative consequences of unprotected sex. It is considered normal for men in
their late teens and 20’s to feel that they are not vulnerable to adverse events (Kreipe,
1985). Thus, young HIV-negative men, aware of the risks of sexually transmitted
infections, may have unprotected sex with the attitude suggesting “it won’t happen to
me”. For the same reasons, HIV-positive men may feel that they will not actually

transmit the virus to their partners. These developmental differences may partly explain

37

our ﬁnding that HIV-positive and HIV-negative MSM were equally as likely to engage in
UAI.

Young men may also engage in unprotected sex because of the tremendous
challenge they face in negotiating safer sex. Our data show that sexual communication is
a problem for men in both serostatus groups with 40% of HIV-negative MSM and 39%
of HIV—positive MSM reporting that they did not talk about safer sex with non-main
sexual partners. Previous research has found that young gay and bisexual men who
engage in unprotected sex are also likely to have poor communication skills with sexual
partners (Hays, Kegeles, & Coates, 1990; Molitor, F acer, & Ruiz, 1999). These young
men may lack the experience and communication skills necessary to resist the pressures
to have unprotected sex. Such barriers to safe sex communication are difﬁcult to
overcome for both HIV-positive and HIV-negative men. The experience of receiving an
HIV-test result does not help young men acquire better communication skills.

Young MSM may also participate in UAI because they are highly sensitive to
peer inﬂuences. In the current study, peer norms for safer sex was protective against
engagement in UAI for both HIV-negative and HIV-positive men. Consistent with Stein
et al. (2005), increased peer norms for safer sex were associated with lower rates of UAI
among the HIV-positive and HIV-negative young men in our sample. The impact of peer
norms on condom use is not surprising. According to Fisher (1988), young men care
about the perceptions of others and will act in ways to retain good standing with their
friends. Fisher advocates that people adhere to group norms in fear of the sanctions that
that can result from nonconformity. While the impact of peer norms on sexual behaviors

can be beneﬁcial, it also presents challenges to HIV/AIDS prevention. For example, the

38

impact of peer norms can be positive in situations where peers hold positive attitudes
towards using condoms. Men a part of these peer groups are more likely to use condoms
if they feel that their peers will judge them unfavorably if they do not use them within
their own sexual encounters (Fisher & Misovich, 1990). On the downside, when men
believe that safer sex is not supported among their peers, they are less likely to use
condoms (Kelly et al., 1995). It is imperative the prevention programming place
particular emphasis on promoting positive norms around safer sex communication and
condom use.

The ﬁnal important similarity detected in our main analyses was relevant to the
helpfulness of prevention among these men. Our ﬁndings indicated that most individuals
in our sample were exposed to at least one form of HIV/AIDS prevention within the
previous three months. The most commonly reported prevention messages were ﬂyers
promoting safer sex (73%) and contact with outreach workers (44%). However, despite
high rates of intervention contact, there was no association between exposure to
prevention and rates of UAI among men in either of the serostatus groups. One possible
explanation for this ﬁnding is that men were most commonly exposed to the types of
interventions that may be least effective. For instance, participation in safer sex
workshops could be more effective in changing sexual risk behaviors than being exposed
to ﬂyers or having contact with outreach workers. It is possible that prevention exposure
did not predict UAI among these men because these individuals are most commonly
reached by forms of prevention that are less potent in their ability to bring about
behavioral change. However, no previous research has directly compared the

effectiveness of each of these types of interventions. Although our measure of prevention

39

exposure was crude, these ﬁndings do suggest that typical prevention activities may be
limited in their ability to reduce rates of unprotected sex among young MSM on a broad
scale. New strategies may be needed in order to reduce rates of unprotected sex with
young MSM.

Differences between HI V -positive and H] V -negative men

Although our ﬁndings identify many similarities among HIV-positive and HIV-
negative men, there were important differences between these groups. Race was an
important factor in predicting UAI for HIV-negative men, but not for those who were
HIV-positive. In addition, HIV-positive men were riskier on some indicators of sexual
behavior. Finally, social support for safer sex was protective against UAI for those men
who were HIV-infected.

Among the HIV-negative MSM, Blacks and Hispanics were the least likely to
have participated in UAI while men in the “other race” category were at the highest risk
for UAI. Previous research conﬁrms that Black MSM are less likely to participate in UAI
when compared to men in other ethnic groups (Harawa et al., 2004). The relationship
between high risk sexual behavior and racial status is paradoxical. While Blacks and
Hispanics report high rates of HIV-infection, they are less likely to report unprotected sex
(CDC, 2006, Valleroy; Mackeller, & Karon, 2000; Lemp, Hirozawa, Givertz, etal.,
1994). Misconceptions held by the public often attribute this discrepancy to injection
drug use. However, IDU was low in our sample (2.3%) and was unrelated to HIV status.
Black and Hispanic MSM who contract HIV in this age group are mostly likely to have

acquired the virus through unprotected sex.

40

In a recent literature review, Millett, Peterson, Wolitski, & Stall (2006) sought to
explain why men in the Black community experience disproportionately higher rates of
HIV in comparison to men in other racial groups. In reviewing the research from the very
beginning of the HIV/AIDS epidemic, the authors concluded that Black MSM do not
engage in higher rates of UAI when compared to MSM in other racial groups. Thus, less
obvious factors are attributed to the spread of HIV among this population. In exploring
these factors, strong scientiﬁc support was found for two of the authors’ hypotheses.
First, Black MSM are more likely to acquire other sexually transmitted diseases which
make them especially vulnerable to contracting HIV. Second, Black MSM are less likely
to know their HIV-status so they may unknowingly expose others to the HIV virus.
Based on these ﬁndings, testing and treatment for sexually transmitted diseases is an
important factor in reducing HIV-infection among black men.

Men who were part of the “other race” category in our study engaged in the
highest rates of UAI. These individuals were largely Asian Americans. The impact of
HIV/AIDS among Asian American men is perplexing. While these men have experienced
the lowest rates of HIV-infections when compared to men in other major ethnic groups in
the United states (CDC, 2005), they report high rates of UAI (Choi, Coates, Catania,
Lew, & Chow, 1995). Researchers have proposed several explanations for this
discrepancy. For instance, It has been suggested that the prevalence of HIV/AIDS among
Asian American MSM is underestimated because these individuals are less likely to get
tested for HIV when compared to men of other ethnic groups (Kahle, Freedman, &
Buskin, 2005; Zaidi et al., 2005). Published reports of HIV/AIDS rates for this population

may also be unreliable because Asian Americans tend to be misclassiﬁed into different

41

racial groups within medical records (Zaidi et al., 2005). Finally, it has been proposed
that Asian American men are less likely to become infected with HIV due to their sexual
mixing patterns. For example, Choi, Operario, Gregorich, & Han (2003) found that Asian
men were most likely to have unprotected sex during sexual encounters with men from
their own ethnic group. Conversely, when having sex with men who were not Asian
American, they tended to practice safer sex.

It should be noted that Asian Americans were underrepresented in our sample of
HIV-positive MSM. With a larger sample of Asian Americans, the relationship between
race and UAI may also be signiﬁcant for HIV-positive men. Therefore, ﬁrture studies
employing larger samples of HIV-positive men from this population is needed in order to
better understand the role of race in sexual risk taking behaviors among HIV-positive
MSM.

In addition to differences related to race, signiﬁcant variations emerged between
the two serostatus group on indicators of sexual risk. Consistent with similar studies, we
found that HIV-positive men were more likely to have engaged in UAI with a greater
number of partners when compared to those who reported to be HIV-negative (Hays et
al., 1997; Valleroy et al., 2000). In addition, we found that HIV-positive MSM were more
likely to have traded sex for food, money, drugs, or a place to stay when compared to
uninfected men. This ﬁnding is also supported in the literature (Hein et al., 1995; Huba et
al., 2000; Pfeifer & Oliver, 1997). Previous research suggests that sex trade is an
economic survival strategy because of its association with homelessness and economic
criminal behaviors among young HIV-infected MSM (Greene et al., 1999). This may also

be true of the men in our sample since HIV—infected men were more likely to be

42

unemployed when compared to their HIV-negative peers. Finally, similar to ﬁndings
reported by Huba et. al. (2000), those who reported an HIV seropositive status were more
likely to have had sex while under the inﬂuence of alcohol or drugs. While we do not
have data that explain this ﬁnding, it has been suggested that HIV-positive men use
substances during sex in order to increase sexual performance and pleasure (Kalichman,
Weinhardt, DiFonzo, Austin, & Luke, 2002) or to facilitate avoidant coping strategies
(Malow, West, Corrigan, Pena, & Cunningham, 1994).

These behaviors pose a serious threat to public health. Not only are the HIV-
positive men in our sample participating in unprotected sex at high rates, they are doing
this with larger numbers of partners when compared to their HIV-negative counterparts.
The prevalence of sex trade among these men is also disturbing since those who engage
in this activity tend to have unprotected sex with large numbers of partners (Raj, Cheng.
Levison, Meli, & Samet, 2006). Finally, the alcohol and drug use occurring during sex
presents a problem in that this behavior is also known to be associated with unprotected
sex (Dolezal, Meyer Bahlburg, Remien, & Petkova, 1997; Hirshﬁeld et al., 2004). HIV-
positive men engaging in these behaviors have the potential to infect thousands of men.

In order to reduce rates of UAI among HIV-infected young men, effective
prevention should address factors that contribute to unprotected sex. Although our data
suggest that the risk and protective factors for UAI are largely similar across the
serostatus groups, one major difference indicated that social support predicted UAI
among HIV-positive men. The role of social support has received less attention in the
HIV prevention literature. Only one previous study has explored the inﬂuence of HIV-

speciﬁc social support among young HIV-positive MSM (Naar-King et al., 2006). The

43

current study is the ﬁrst to compare HIV-positive to HIV-negative MSM using such a
measure of social support. Our ﬁndings indicate that HIV-positive men who reported
increased levels of social support for safer sex were less likely to engage in UAI. Thus,
having individuals with whom one can discuss issues related to safer sex was especially
protective for these men.

Increased social support may play a role in sexual behaviors for a number of
reasons. First, it is possible that HIV-positive men who have social support are more
likely to disclose their HIV-serostatus to sexual partners (Kalichman, 2003). Research
shows that HIV-infected men who inform their sexual partners of their serostatus are less
likely to have unprotected sex (Klitzman, 2007). Second, men who have social support
for safer sex may receive informational support concerning condom use and strategies for
negotiating safer sex. In turn, these individuals experience increased self-efﬁcacy. For
example, Naar-King et al. (2006) found that social support leads to greater self-efﬁcacy,
which resulted in safer sexual behaviors among HIV-infected young men.

Finally, having family and friends involved in the sexual decision making of these
men may inﬂuence HIV-positive men to take the personal responsibility to protect the
health of their sexual partners. In one study, Parsons, Halkitis, Wolitski, Gomez, & the
Seropsoitive Urban Men’s Study Team (2003) found that 25% of adult HIV-positive men
saw themselves as having little personal responsibility for infecting others. In another
recent article, it was found that having the belief that it is one’s partner’s responsibility to
protect himself, rather than a personal responsibility to protect him, is associated with
high sexual risk (Wolitski, Flores, O'Leary, Bimbi, & Gomez, 2007). Thus, social support

may deter some men from having unprotected sex because it makes personal

44

responsibility more salient. When these young men go to their friends and family
members for safer sex support, they may become more accountable for their actions.
Thus, social support may reinforce these men to take responsibility for protecting their
partners by avoiding UAI.

Future research should explore the effect of HIV-risk speciﬁc social support on
disclosure, self-efﬁcacy, personal responsibility and other factors that may mediate the
relationship between social support and high risk sex.

Limitations

There are some limitations that Should be considered when interpreting the results
of this study. First, these results are based on self-reports. It is possible that some
participants may have given us inaccurate information about their HIV-status due to fears
of being stigmatized. Similarly, fears of stigmatization, as well as biases associated with
social desirability may have led some individuals to have misreported information
concerning their sexual behaviors. However, the likelihood of misreporting should have
been reduced since participation in this study was anonymous. In addition, our results are
consistent with other studies using similar sampling and methodological procedures.

Second, we cannot assess the causal impact that an HIV diagnosis has on sexual
behaviors. Since we do not have longitudinal data pertaining to the sexual behaviors of
men before and after their diagnosis, it is not possible to determine the exact impact of
serostatus on sexual behaviors. The HIV-infected men could have become more or less
risky in their behaviors post-diagnosis. For example, previous longitudinal research with
HIV-positive adults indicates that individuals decrease their sexual risk behaviors

immediately following HIV-diagnosis, but risky sexual behaviors tend to increase again

45

over time (Colfax et al., 2002; Di-Franceisco, Pinkerton, Dyatlov, & Swain, 2005).
Future longitudinal research exploring the trajectory of sexual risk behaviors pre and post
HIV diagnosis with young MSM is needed.

Third, because the participants in this study were not subjected to serological
testing, their actual HIV status was not conﬁrmed. Consequently, it is likely that the
prevalence of HIV in this sample is underestimated. Previous researchers have reported
rates of HIV infection to be between 7-19% (Hays et al., 1997; Hein et al., 1995;
Valleroy et al., 2000). However, our self-reported rates of HIV infection are consistent
with Valleroy et. al., (2000) who used a Similar sample of young MSM and reported that
1.3% of the men perceived themselves to be HIV-positive (the self—reported rate of HIV
was 2% in the current study).

Another potential limitation of this study is that we do not have data on the sexual
partners of the HIV-positive men who engaged in UAI. It could be argued that men
having sex with seroconcordant partners are not at risk. However, unprotected sex is a
serious health threat regardless of whether or not it occurs between sexual partners with
the same HIV-status since these men can contract other sexually transmitted diseases.
Moreover, many people living with HIV/AIDS already have compromised immune
systems, such infections can therefore be especially harmful. In addition, although it is
less common, UAI can pose a threat to HIV-infected individuals in that they could
become re-infected with a different strain of the HIV-virus (Jost et al., 2002; Smith et al.,
2004)

Finally, it should be noted that two of the ﬁndings reported here were only

marginally statistically signiﬁcant. Speciﬁcally, the relationships between sex trade and

46

serostatus, as well as substance use and serostatus, were signiﬁcant at the p = .056 and p
= .06 respectively. Future research employing larger samples of young HIV-positive
MSM is needed in order to conﬁrm these results.

Implicationsfor Prevention

The single event of receiving an HIV-positive diagnosis is not enough to change
the sexual behaviors of many young MSM. In response, prevention initiatives have
slowly expanded to include speciﬁc programming for HIV-positive men in recent years.
These evidence-based programs have made some advances in helping HIV-positive
individuals adopt safer sexual practices (Crepaz et al., 2006). However, few of these
prevention programs have speciﬁcally targeted younger populations of HIV-positive
MSM aged 25 and younger. Existing interventions typically approach prevention under
the assumption that similar factors predict risky sexual behaviors among HIV-positive
and HIV-negative individuals. Although our results demonstrated that there are both
similarities and differences between HIV-negative and HIV-positive young MSM in
terms of sexual risk, there were also important differences between the two groups which
suggest that serostatus should be considered during the development and implementation
of prevention programming.

Since peer norms for safer sex were important in reducing UAI for men in both
serostatus groups, effective prevention programming should pay special attention to
assisting all young MSM in forming relationships with peers who endorse the use of
condoms and also use them in their own encounters. One intervention that has focused on
changing peer norms among young MSM is the Mpowerment Project. Mpowerment is an

evidence-based, community level program targeting young MSM. The peer-based

47

component of this program attempts to mobilize young MSM to talk to their friends about
safer sex (Hays, Rebchook, & Kegeles, 2003; Kegeles, Hays, Pollack, & Coates, 1999).
Interventions such as the Mpowerment Project are guided by the diffusion of innovation
theory (Rogers, 1995). Other interventions based on this theory have also been effective
in reducing risky sex by changing peer norms through the recruitment and training of
opinion leaders who serve as safer sex endorsers within their social networks.
Differences between HIV-positive and HIV-negative men that emerged in this
study suggest that HIV-prevention programs for young MSM may be enhanced by
addressing serostatus speciﬁc issues. For instance, a large majority of young MSM living
with HIV/AIDS are African American or Hispanic suggesting that HIV-prevention
programs should be sensitive to cultural issues around sex and sexuality. In addition,
prevention for HIV-positive young MSM should place special emphasis on helping
young MSM expand their networks of social support to include family and friends with
whom they can discuss issues related to safer sex. Although there are family-level
interventions that typically aim to increase communication in regards to HIV and other
sexually transmitted diseases (DiClemente, Salazar, & Crosby, 2007), these types of

programs tend to target adolescent populations rather that young HIV-positive MSM.

48

Conclusions
Unfortunately, the rate of new HIV diagnoses among young MSM aged 25 and
under is growing rapidly. With an increasing number of young men who are living with
HIV/AIDS, the prevention needs of this population must be addressed. Future directions

in research should explore developmental, cultural, and social needs of this population.

49

APPENDIX A

QUANTITATIVE
INTERVIEW (QTI) CITY
Study - October 26,1998

The QTI is to be completed only for consenting eligible YMSM as deﬁned by BSI.

Start time: : am / pm
hour minute (circle one)

A. Demographics

A 1. What is your date of birth? / /

mm dd W

A2. What zip code do you live in?

 

A3. How many years of school have you completed?
years (examples: "ﬁnished high school"=12, "bachelors degree"=16)

A4. Are you currently enrolled in school, or were you enrolled last year? (Prompt:
Full or part time?,) (Note: By "last year," we mean the academic year that started
in the fall of 1998 and ended in the spring of 1999.)
1) Yes, Full time student
2) Yes, Part time student
3) No

A5. How would you describe your current work situation?
1) Full time paid job (at least 30 hourS/tlveek)
2) Part time paid job (less than 30 hours/week)
3) Not working because of disability
4) Not employed, looking for work
5) Not employed, not looking for work
6) On unemployment (receiving unemployment beneﬁts)
7) Other (Specify: )
8) Refused

A6. Were you born in the United States?
1) Yes

A6a. In what city and state were you born?

City , State (Skip to question A9)
2) No

50

A6b. In what country were you born? (Country
And in what city? (City )

 

(These items are optional, site-speciﬁc)

A 7. How old were you when you ﬁrst came to live in the mainland United

 

States?
_ years
A8. In the last year, have you been back to (insert country of
birth)?
1) Yes
2) No
A9. In general, what language do you read and speak?
1) Spanish 1) English only
2) English or 2) English primarily
3) Both 3) Both equally
4) Other language primarily
5) Other language only

6) Declined to answer

A 10. What language do you usually speak at home?

1) Spanish 1) English only

2) English or 2) English primarily
3) Both 3) Both equal/y

4) Other language primarily

5) Other language only

6) Declined to answer

A11. In what language do you usually think?

1) Spanish 1) English only
2) English or 2) English primarily
3) Both 3) Both equally

4) Other language primarily
5) Other language only
6) Declined to answer

A 12. What language do you usually speak with your friends?

1) Spanish 1) English only

2) English or 2) English primarily

3) Both 3) Both equally
4) Other language primarily
5) Other language only

6) Declined to answer

51

A 13. Do you consider yourself to be: (Interviewer: read options and check only
one)

1) Gay (Homosexual)

2) Bisexual

3) Straight (Heterosexual)

4) Undecided / Don 't Know

A13. Do you consider yourself to be: (Interviewer: read options and check only one)
1) Gay (Homosexual)
2) Bisexual
3) Straight (Heterosexual)
4) Undecided / Don ’t Know
5) Other
6) Refused

52

B. Female Partners

Now I will be asking you some personal questions about sex. Remember your
answers are confidential. We're not interested in the names of people.

B1.

82.

33.

Thinking back over the past three months, have you had any kind of sexual
contact with a woman? By sexual contact, we mean vaginal, oral, or anal sex,
or any other physical contact leading to orgasm.

1) Yes

2) No (Skip to Section C: Male Partners in Past Three Months)

The last, most recent time you had any kind of sexual contact with her, did
you have vaginal sex?

1) Yes

2) No (Skip to question BB)

B2a. While you were having vaginal sex that time, did you use a condom
from start to ﬁnish? (If necessary: eithera male condom or a female
condom, )

1) Yes
2) No

The last, most recent time you had any kind of sexual contact with her, did
you have anal sex?

1) Yes

2) No (Skip to Section C: Male Partners in Past Three Months)

B3a. While you were having anal sex that time, did you use a condom from
start to ﬁnish?
1) Yes
2) No

53

C. Male Partners in Past Three Months

Next I'd like to ask you a few questions about any sex you 've had with male
partners in the past three months - that is, since "exact date three months ago", or
about "familiar date or holiday three months ago." When lsay "men”, lmean male
partners of any age.

C1. In the past three months, how many men have you had any sexual contact
with? By sexual contact, I mean oral or anal sex, or any other physical
contact leading to orgasm.

# men (If CI is m then verify: So you have not had any kind of
sex with
male partners in the past three months? and skip to section
D: Sex with Main Male Partner)

Nowl want to ask some questions about anal sex. When I say "anal sex", I
mean penetrative anal intercourse.

(lf C1 above is % then skip to question C4)

(If C1 above is more than one. then proceed to question CZ)

C2. (If more than one male partner) So of those (insert number from question C1)
male partners you just mentioned, with how many different men have you
had any anal sex in the past three months? No matter whether you were the
inserting partner or the receptive partner, and no matter whether you used
condoms or not.

# men (If CZ is zero, then verify: So you have not had anal sex with
any of

your male partners in the past three months? and skip to section D)
(If C2 is oﬂ, then skip to question C4)
(If CZ is more than one, then proceed to question CB)
C3. And of those (insert number from question CZ) male partners for anal sex,
with how many different men have you had any anal sex without a condom
in the past three months?

# men
Nowl want to change the topic from the number of partners to the number of

different times you 've had anal sex with and without condoms in the past three
months.

54

C4.

C5.

C6.

C7.

In the past three months, how many times have you had anal sex with men (or
this man), where you were the inserting partner, and you grg use a condom?

# times

In the past three months, how many times have you had anal sex with men (or
this man), where you were the inserting partner, and you did not use a
condom?

# times

In the past three months, how many times have you had anal sex where you
were the receptive partner, and your partner g_ir_l use a condom?

# times

In the past three months, how many times have you had anal sex where you
were the receptive partner, and your partner did not use a condom?

# times

55

D. Sex with Main Male Partner

Now I want to ask you some questions about the last time you had sexual contact
with a man who you consider to be your main male partner, like a lover or a
boyfriend. Remember your answers are conﬁdential.

D1. In the past M, have you had any kind of sexual contact with a man
who you consider to be your main male partner? (repeat deﬁnition of sex
(from page 4, C1))
1) Yes
2) No (Skip to Section E: Sex with Non-Main Male Partner)

02. About how long ago was the last time you had sexual contact with a main
male partner?
days / weeks / months ago / today (Circle one and write number in blank)

DB. The last, most recent time you had any kind of sexual contact with him, did
you have §n_al sex where you were the inserting partner?
1) Yes
2) No (Skip to question D4)

03a. While you were having anal sex that time, did you use a condom from
start to ﬁnish?
1) Yes
2) No

D4. And the last, most recent time you had any kind of sexual contact with him,
did you have plat sex where you were the receptive partner?
1) Yes
2) No (Skip to question D5)

D4a. While you were having anal sex that time, did he use a condom from
start to ﬁnish?
1) Yes
2) No

D5. Were you high on drugs or alcohol the last time you had sex with him?
1) Yes
2) No

D6. The last time you had sexual contact with him, did you have an idea what his
HIV status might be?
1) Yes, I did
2) No, ldid not (Skip to Section E: Sex with Non-Main Male Partner)

D7. Could you tell me what you thought his status was at that time?
1) I thought it was Positive
2) I thought it was Negative
3) Refused (Skip to Section E: Sex with Non-Main Male Partner)

56

DB. Why did you think that he was (response from D7)7
(Do not read options aloud. Instead, prompt for response and categorize.)
1) He told me explicitly
Z) I guessed based on other things he has talked about
3) Someone else told me
4) By his appearance
5) We got tested together
6) I saw his results
7) I went with him to get his results

 

8) Other (Specify: )

09. Have the two of you ever talked about using condoms or practicing safer sex?
1) Yes
2) No

57

LSex with Non-Main Male Partner

Nowl want to ask you some questions about the last time you had sexual contact
wiﬁr a man who you did not consider to be your main male partner. Remember your
answers are conﬁderrﬁal.

E1.

E2.

E3.

E4.

E5.

E6.

E7.

In ﬁre past reg; have you had any kind of sexual contact with a man who was
not your main male partner? (repeat deﬁnition of sex (from page 4, C1 ))

1) Yes

2) No (Skip to Section F: Other Risk History)

About how long ago was the last time you had sexual contact with a non-
main male partner?
days / weeks / months ago / today (Circle one and write number in blank)

The last, most recent time you had any kind of sexual contact with him, did
you have ,aLaI sex wiﬁr you as ﬁre inserting partner?

1) Yes

2) No (Skip to question E4)

533. While you were having anal sex ﬁrat ﬁme, did you use a condom from
start to ﬁnish?
1) Yes
2) No

And ﬁre last, most recent ﬁme you had any kind of sexual contact wiﬁr him,
did you have pn_al sex wiﬁr you as ﬁre recppﬁve partner?

1) Yes

2) No (Skip to question E5)

E 4a. While you were having anal sex ﬁrat time, did he use a condom from
start to ﬁnish?
1) Yes
2) No

And ﬁre last, most recent ﬁme you had m kind of sexual contact wiﬁr him,
did you communicate wiﬁr him in any way about using a condom when you
had sex ﬁrat ﬁme?

1) Yes

2) No

Were you high on drugs or alcohol ﬁre last ﬁme you had sex with him?
1) Yes
2) No

The last ﬁme you had sexual contact wiﬁr him, did you have an idea what
his HIV status might be?

1) Yes, I did

2) No, I did not (Skip to Section F: Other Risk History)

58

E8. Could you tell me what you thought his status was at that time?
1) I thought it was Positive
2) I thought it was Negative
3) Refused (Skip to Section F: Other Risk History)

E9. Why did you think that he was (response from E8)7
(Do not read options aloud. Instead, prompt for response and categorize.)

1) He told me explicitly
2) I guessed based on other things he talked about
3) Someone else told me
4) By his appearance
5) We got tested together
6) I saw his results
7) I went with him to get his results
8) Other (Specify: )

59

F. Other Risk Histogy

Now I have a couple of oﬁrer quesﬁons to ask before I go on to a completely
different topic.

F 1. In ﬁre last year, has anyone given you money, drugs, or anyﬁring valuable
(like food, or a place to stay) to have sex?
1) Yes
2) No

F2. Have you injected non-prescripﬁon drugs in ﬁre last year?

1) Yes
2) No

60

G. Peer Norms about Condoms

People have different attitudes toward condoms, and we 'd like to get a sense about
what ﬁre people you know ﬁrink about condoms. How much do you agree or
disagree wiﬁr each of the following statements? (Show respondent response card)

G1. Most of my friends ﬁrink you should always use condoms when having anal

sex with a new partner.
1) Agree Strongly
2) Agree Somewhat
3) Disagree Somewhat
4) Disagree Strongly
5) Don 't know/No response

62. Most of my friends are using condoms ﬁrese days when they have anal sex.
1) Agree Strongly
2) Agree Somewhat
3) Disagree Somewhat
4) Disagree Strongly
5) Don't know/No response

G3. Most of my friends think you should use condoms whenever you have
anal sex, including wiﬁr a primary sex partner.
1) Agree Strongly
2) Agree Somewhat
3) Disagree Somewhat
4) Disagree Strongly
5) Don‘t know/No response

G4. Most of my friends use condoms when they have anal sex with a new partner.
1) Agree Strongly
2) Agree Somewhat
3) Disagree Somewhat
4) Disagree Strongly
5) Don 't know/No response

61

H. Social Support

We all look to oﬁrers for ﬁiendship, to talk wiﬁr, and to get oﬁrer kinds of help at one
ﬁme or anoﬁrer. In ﬁre last 3 monﬁrs, how often has each of ﬁre following kinds of
help been available to you, if you needed it? (Show respondent response card)

HI Someone you can share concern about AIDS wiﬁr.
1) Never or none of the time
2) Rarely or little of the time
3) Sometimes
4) Most of the time
5) Always or all of the time

H2. Someone you could talk to about your sexual fantasies.
1) Never or none of the time
2) Rarely or little of the time
3) Sometimes
4) Most of the time
5) Always or all of the time

H3. Someone you can talk to about safer sex.
1) Never or none of the time
2) Rarely or little of the time
3) Sometimes
4) Most of the time
5) Always or all of the time

H4. Someone you could tell if you were having unsafe sex.
1) Never or none of the time
2) Rarely or little of the time
3) Sometimes
4) Most of the time
5) Always or all of the time

62

I. Knowledgg and Attitudes about Antiretrovirals

Okay, now I 'd like to ask you a few questions on anoﬁrer topic.

1 1. Have you heard about using a combinaﬁon of drugs to treat HIV/AIDS,
refened to someﬁmes as 'combinaﬁon drug cocktails? (Someﬁmes ﬁrey
include dnrgs called protease inhibitors.)

1) Yes
2) No (Skip to Section J: Exposure to interventions)
3) Not Sure

Tell me how much you agree or disagree wiﬁr ﬁre following statements:

12 Because ofﬁrese drugs, HIVisaIessserious ﬁrreatﬁranitusedto be.
1) Agree Strongly
2) Agree Somewhat
3) Disagree Somewhat
4) Disagree Strongly
5) Don 't know/No response

13 Ipractice safe sexless often now because new medical treatments for HMAIDS
have
come along.
1) Agree Strongly
2) Agree Somewhat
3) Disagree Somewhat
4) Disagree Strongly
5) Don 't know/No response

63

J. Exposure to Interventions

I 'd like to ask about ﬁre different kinds of AIDS prevention activities
you may have encountered. When I say "AIDS prevenﬁon ", I mean

using condoms or avoiding unprotected sex.

 

(Insert local project's name in blanks)

Yes
No

How
Many

#From
Project

 

In the past six months, have you participated
in any workshops about AIDS prevention?
How many workshops? How many of these
were hosted by?

 

In the past six months, have you
attended any social events or parties that
included AIDS prevention activities? How
many events? How many of these were
hosted by?

 

In the past six months, have any
educators or outreach workers talked to
you about AIDS prevention? How many
times? How many times were these
people from?

 

In the past six months, have you seen any
ﬂyers or pamphlets about AIDS prevention?
How many different ﬂyers? How many of
these flyers were from?

 

(Show samples)

 

In the past six months, have you seen any
ﬂyers or pamphlets with information on
workshops or social events? How many
different ﬂyers? How many of these ﬂyers
were from?

 

(Show samples)

 

Have you seen or heard anything else
about?

 

 

(Example: radio, newspaper, TV)

 

 

 

 

64

 

K. Wrap-up

I've got just a few more questions to ask. These won't take long at all and then we 're
done.

K 1. We'd like to know whether people cany condoms with ﬁrem. Are you

canying a condom now?
1) Yes
2) No (Skip to question K2)

K 1 a. Would you show it to me?
1) Condom seen, marked with CITY project logo [for use only after
Intervention begins]
2) Condom seen, not from CITY project
3) Condom nearby in coat, car, etc.

K2. Have you been interviewed with ﬁris survey before?
1) Yes
2) No (Skip to question K6)
3) DK / NS (Skip to question K6)

K3. How many times? #

K4. When was that/ ﬁre last time? That is, how long ago?

(Record answer as given by respondent, then categorize as 1, Z, 3, 4, or 5 below.)

Response:
1) Less than 1 month ago
2) 1 to 2 months ago
3) 3 to 5 months ago
4) 6 months to a year ago
5) more than a year ago
6) DK / NS

K5. Where were you interviewed ﬁrat last time?

K6. In order to ﬁnd oﬁrer men like yourself to do ﬁris survey, we would like to
know where you and your friends might hang out and how often you go
there. (Hand respondent palm card with list of venues in that city.) Can you tell
me the total number of times you have been to any of these places in ﬁre past
two weeks?

times

65

K7. Are ﬁrere any other places not on ﬁris list ﬁrat you or your friends would go
to?

Name of place Best days Time of day

K8. Can you name a place where you would feel comfortable going for an HIV
test?

1) Yes (Specify: )

2) No

K9. Have you ever been tested for HIV?
1) Yes
2) No (Skip to question K12)
3) DK / NS (Skip to question K12)
4) Refused (Skip to question K12)

K 10. About how long ago was the last ﬁme you were tested for HIV?
_ days / weeks / months / years ago (Circle one and write number in blank)

K11. Would you be willing to tell us the result of the last test?
1) Positive
2) Negative
3) lndetenninate (Verify: So the doctor or clinic said the result was
indeterminate?
4) Did not go back to get results
5) Don't know (Prompt: Did you go back to get the results ?;
6) Refused

K12. Now that the survey is completed, would you like to go back and make
changes to any of your responses?
1) Yes (record changes next to original response)
2) No

End time: : am / pm
hour minute (circle one)

That was the last question. You 've been very helpful. Thank you for your time and

patience. Do you have any questions about what we 've talked about ﬁrat you would
like to ask me?

66

Correct signiﬁcant misconceptions about AIDS transmission/prevention and
make referrals as appropriate

Interviewer comments:

Do you have any reason to question the tnrthfulness of the respondent’s answers?
Yes (discuss this with your supervisor and ﬁll out the appropriate form)
No

Do you think, for any reason, that the respondent is ineligible for the study?
Yes (discuss this with your supervisor and ﬁll out the appropriate form)
No

67

APPENDIX B

The CITY data were entered and cleaned using state-Of-the-art procedures for
insuring data quality. First, all data were double entered at the sites on a custom-designed
data entry template that was designed to minimize data entry errors. After site data entry
staff had addressed any data discrepancies identiﬁed by double entry comparison, data
were forwarded on a monthly basis to a professional data coordinating center, which ran
analyses on data inconsistencies and illogical patterns in the data. Data inconsistencies
were things such as discrepancies between a respondent’s age and birth date. Illogical
reporting were things such as stating one had no casual partners and then later reporting
sexual acts with a casual partner. Once discrepancy and illogic reports were produced,
these were returned to the sites who sought to correct problems that resulted from data
input errors. Data for which problems could not be resolved were removed from the
analytic sample if the nature of the discrepancy called into question the validity of the

data or concerned study eligibility criteria.

68

APPENDIX C

Table 1

Cities and Racial Strata UV =83 75)

 

 

Community N

African American

Atlanta 812

Birmingham 585

Chicago 542
Hispanic/Latino

LA/Orange County 579

LA/San Gabriel Valley 560

NY/Jackson Heights/Corona 775

NY/Washington Heights/South Bronx 756
Asian/Paciﬁc Islander

San Diego 355

Seattle 348
Mixed Race

Detroit 816

Milwaukee 816

Minneapolis 732

LA/West Hollywood 731

 

69

Table 2

Descriptive Statistics of Study Variables among HIV-negative and HIV-positive young
MSM

 

 

 

 

HIV- negative HIV-positive Total
Predictor Variables 0
% N % N /0 N x2

Race 22.84**

Hispanic 34.2 2811 34.5 59 34.2 2870

African American 27.8 2283 40.9 70 28.1 2353

White 22.6 1858 10.5 18 22.4 1876

Other 15.4 1261 14 24 15.3 1285
Education 3.88

0-6 Years .6 52 1.2 2 .6 54

Some H.S 10.0 819 14.1 24 10.1 843

H.S Grad 35.1 2874 35.3 60 35 2934

Some College 40 3282 36.5 62 40 3344

College Grad 14.3 1169 12.9 22 14.2 1191
Employment 6.23**

Not Working 14.4 1185 21.6 37 14.6 1222

Working 85.6 7020 78.4 134 85.4 7154
Sexual Orientation 5 .8 1

Gay 74.1 6074 80.1 137 74.2 621 l

Bisexual 20 1643 14 24 19.9 1667

Straight .5 38 - 0 .5 38

Other 5.4 441 5.8 10 5.4 451

Mean SD Mean SD Mean SD t-test

Age 21.52 2.26 22.18 2.4 21.53 2.26 -3.78**

 

*p<.05, **p.<.01

70

Table 3

High Risk Sexual Behaviors of H] V -negative young MSM versus those who are HIV -
positive

 

HIV- negative HIV-positive
Predictor Variables

 

% N % N x2
Within Past Year
Traded Sex (N=8375) 3.66”r
Yes 10.4 854 15.2 26
No 89.6 7950 84.8 145
Past 3 Months
UAI with Men (N=753 8) 6.23
Yes 34 2481 39.6 61
No 66.4 4903 60.4 93
UAI or UVI with women (1003) ***
Yes 18.2 181 27.3 3
NO 81.8 81 1 72.7 8
Last Encounter
UAI with main partners (N=6152) .03
Yes 25 1508 24.4 30
No 75 4521 75.6 93
UAI with non-main partners (N=5969) 2.58
Yes 9.6 559 14.1 18
NO 90.4 5282 85.9 110
Sex while high w/ main partners 3 46’
(N=6167) '
Yes 17.6 1066 24.4 30
No 82.4 4978 75.6 93
Sex while high w/ non-main partners 1 98
(N=5990) '
Yes 27.6 1619 33.3 43
No 72.4 4246 66.7 86
Knowledge of main partner’s HIV 02
status (N=6176) '
Yes 84.2 5098 83.7 103
No 15.8 955 16.3 20

71

Knowledge of non-main partner’s HIV
status (N=5984)

Yes

NO

Talk about safer sex with main partner
01:5 797)

Yes

No

Talk about safer sex with non-main
partner (N=5980)

Yes

No

Number of male partners without a
condom in past 3 months (N=8295)

 

 

1.50
41 2399 35.7 46
59 3456 64.3 83
.11
84.9 4828 84.1 95
15 855 15.9 18
.05
60.2 3526 61.2 79
39.8 2329 38.8 50
Mean SD Mean SD t-test
.36 .61 .49 .72 -2.24*

 

72

_o.VQ..:.. .mo.VQ*
.monHZ .802

 

 

:2. :8. ._.8.- 1.8.- 1.8.- 8. 88.- 5.- :8.- ..8. .8. S.- .....8. 8285335.:
- ooo. moor **mor foot .36. tumor too: _oo. **oo.. taco—- **oor moor xom nova-rd—
- 88.- .8.- ..E; 8. 8.- 8. 8. :8. :8.- ..._.S.- 8. :3 852
- :8. 88.- :8. So.- ..8. 8. S. .8.- ooo. :8.- 8:55;:
- :8. 8.- 88.- :8. 88. :8. 1.8.- :8. 8.- EERESRS
- 8.- .88. :8.- .....8.- t. _ _.- :8. 3.8. 5. 852 cos .8
- .18.- 8.- :8. 2.8.- So. :8. 8. seem >5 .8
- :2. 88. :8. 8. :8.- ..s:. sneeze-n:
- _o. i. _o_. **vmr **ho. 3.3”. comuaoaom .o
- :8.- .....8.- :8.- S. 82 .25 .m
- **om.- stem: ..&or 353 .v
- to..- :8. 8:3 .m
- :8.- esm .N
- 8a...
2 2 : 2 a w e e m 4 m m _

 

833.88 REE-w. the: Losses-So: weotSstob 83.83%

v 03m?

73

.ocl do» H o Boob mg» .35 En die-ES “concave o5 .bzﬁﬁeauBE mo 8092. of cor-— ..

 

 

 

 

 

 

 

Eva... .88...
. . 8.25m
3 vo m oo—uoﬁowm Z
8a: 2.42 82:85 85 NW
..... a 8 8 m 28$ :6 Em”
8s 8. 8. 8._ 2..- a: 00335
:8 2. S. 8. 8.- am 865
ccmwao>en~
8.2 mm. 8.. 8. 8.- e osmoaxm
a? a: t._ 3.8.8 2. 88% seem
8; A: m: .88. S. eaoz as.
8.4 8. N: R; E. 88 S. :1 6. 4m. BEES
N2 9.. 8. m2 8.. w: a. 8. N3 8.. 898:3
Em om. em. 8. 8.- Ba 8. 8. 8. 8.- see»
a: NS. 83 E. R. was 3.. e: 8. 2. seam
Se 8. a: 2: 8. 8.4 an. a? E. 2.. 8:3
RE 8. of mm. 8. w: 8. :2 8. 8. 8<
8&3 8.53 comm 2a? m 8&3 base-N comm 255 m 832.252
88 see
6 .88 D .88
m meow _ meow
site

3%: M2368 ssESQ-xzt MESS $380385 Nets 32986-385 M38338 Embers BBS-Doom £3ch Nstsmssﬁ.

m Eon-w

74

 

 

.o:no do» H o 380 mm?» ._<3 58 638:; Eoocomow 2t .b:£82&8£ oo 3095a 05 “on .,
EVE; .mOVQ...

 

 

75

 

 

 

 

 

. . 896m
no mo m 388—3me
8.9.8 8.83 82:35 ms N-
. . ca
.1. Amv mom Sum **Gv mm #2 oazcm EU Etc—Z
wow. wow. no. mm.m o_.- 83 oozﬁmosm
8v. 8.... a. S. 8... 5m 865
. . . . . . 5:555
Nmo mmo on oo _o 8 Bzmooxm
mg. mil. R. oo. moo. tooazm Eoom
com. com. m _ ._ 13.53 mo. 8802 sum
_o._ on. 34 CM- 2. EA o._ ww. *mw.m 3. Eoﬁhoiﬁm
mo; mo. oo. mo. ooor no; vo. _o._ mo. woo. coswosom
54 E. o 2 Sum m _. vmg No. ow. no. co. BE?
om. mv. mo; .13.? mo. nmd E..— S. 1%ko K. mom—m
mm. mm. no; 3.2.?” xv. a.“ :1 No. 3.5.? om. 0:25
mo; o._ mo. mod Nor o: mo. mo; End mor ow<
ERAS Sic-N 23% 255 m EASE $.55 ovum 335 m moEmEm>
£60 £85
5 $3 5 $3
m @on M 3%
an? ”>2

3%: MESA 9:2»?283‘ MESS ESESSEN BSx 3292335 MEBNBEK ﬂwbgw ncNEEMmm 9.28ch 32:33.0

0 @33-

REFERENCES

Bailey, S. L., Camlin, C. S., & Ennett, S. T. (1998). Substance use and risky sexual
behavior among homeless and runaway youth. Journal of Adolescent Health,
23(6), 378-388.

Beck-Sague. (2004). HIV/AIDS. Philadelphia: Chelsea House Publishers.

CDC. (1998). 1998 guidelines for treatment of sexually transmitted diseases. Morbidity
and Mortality Weekly Reports, 27(1), 1-118.

CDC. (2001). Revised guidelines for HIV counseling, testing, and referral. MMWR,
50(RR-19), 1-57.

CDC. (2003a). From the Centers for Disease Control and Prevention. Advancing HIV
prevention: new strategies for a changing epidemic -- United States, 2003. JAMA:
the journal of the American Medical Association 289(19), 2493-2495.

CDC. (2005a). Trends in HIV/AIDS Diagnoses - 33 states, 2001-2004. Morbidity and
Mortality Weekly Report, 54(45), 1129-1153.

CDC. (2005b). HIV Prevention in the Third Decade. Centers for Disease Control and
Prevention.

CDC. (2006). Best-Evidence Interventions. Retrieved January 20, 2007, from
http://www.cdc.gov/hiv/topics/research/prs/best-evidence-intervention.htm

CDC. (2007). HIV/AIDS Surveillance Report: Cases of H] V Infection and AIDS in the
United States and Dependent Areas, 2005. Atlanta: Centers for Disease Control &
Prevention.

Choi, K. H., Coates, T. J., Catania, J. A., Lew, S., & Chow, P. (1995). High HIV risk
among gay and Asian and Paciﬁc Islander men in San Francisco. Aids, 9(3), 306-
308.

Cohen, S. (1988). Psychosocial models of the role of social support in the etiology of
physical disease. Health Psychology, 7(3), 269-297.

Colfax, G. N., Buchbinder, S. P., Comelisse, P. G. A., Vittinghoff, E., Mayer, K., &
Celum, C. (2002). Sexual risk behaviors and implications for secondary HIV
transmission during and after HIV seroconversion. Aids, 16(11), 1529-1535.

Crepaz, N., & Marks, G. (2002). Towards an understanding of sexual risk behavior in

people living with HIV: A review of social, psychological, and medical ﬁndings.
Aids, 16(2), 135-149.

76

Crepaz, N., Lyles, C. M., Wolitski, R. J., Passin, W. F., Rama, S. M., Herbst, J. H., et al.
(2006). Do prevention interventions reduce HIV risk behaviors among people
living with HIV? A meta-analytic review of controlled trials. AIDS, 20(2), 143-
157.

Denning, P. H., & Campsmith, M. L. (2005). Unprotected Anal Intercourse Among HIV-
Positive Men Who Have a Steady Male Sex Partner With Negative or Unknown
HIV Serostatus. American Journal of Public Health, 95(1), 152—158.

Deren, S., Beardsley, M., Coyle, 8., & Singer, M. (1998). HIV serostatus and risk
behaviors in a multisite sample of drug users. Journal of Psychoactive Drugs,
30(3), 239-245.

Di-Franceisco, W., Pinkerton, S.-D., Dyatlov, R.-V., & Swain, G.-R. (2005). Evidence of
a Brief Surge in Safer Sex Practices After HIV Testing Among a Sample of High-

Risk Men and Women. Journal of Acquired Immune Deﬁciency Syndromes,
39(5), 606-612.

DiClemente, R. J ., Salazar, L. F., & Crosby, R. A. (2007). A review of STD/HIV
preventive interventions for adolescents: sustaining effects using an ecological
approach. Journal of Pediatric Psychology,, 32(8), 888-906.

Dolezal, C., Meyer-Bahlburg, H. F. L., Remien, R. H., & Petkova, E. (1997). Substance
Use During Sex and Sensation Seeking as Predictors of Sexual Risk Behavior
Among HIV+ and HIV- Gay Men. AIDS and Behavior, 1(1), 19-28.

Field, A. P. (2005). Discovering Statistics Using SPSS (2nd ed.). London: Sage
Publications Inc.

Fisher, J. D. (1988). Possible effects of reference group-based social inﬂuence on AIDS-
risk behavior and AIDS-prevention. American Psychologist, 43(1 1), 914-920.

Fisher, J. D. & Misovich, S. (1990). Social inﬂuence and AIDS-preventive behavior. New
York, NY, US: Plenum Press.

George, N., Green, J ., & Murphy, S. (1998). Sexually transmitted disease rates before and
after HIV testing. International Journal of STD & AIDS, 9(5), 291-293.

Glynn, M., & Rhodes, P. (2005, June). Estimated H] V prevalence in the United States at
the end of 2003. Paper presented at the National HIV Prevention Conference,
Atlanta.

Goldbaum, G., Perdue, T., Wolitski, R., Rietmeijer, C., Hedrich, A., Wood, R., et al.
(1998). Differences in risk behavior and sources of AIDS information among gay,

bisexual, and straight-identiﬁed men who have sex with men. AIDS and Behavior,
2(1), 13-21.

77

Greene, J. M., Ennett, S. T., & Ringwalt, C. L. (1999). Prevalence and correlates of
survival sex among runaway and homeless youth. American Journal of Public
Health, 89(9), 1406-1409.

Halcaon, L. L., & Lifson, A. R. (2004). Prevalence and Predictors of Sexual Risks
Among Homeless Youth. Journal of Youth and Adolescence, 33(1), 71-80.

Haley, N., Roy, E., Leclerc, P., Boudreau, J. F., & Boivin, J. F. (2004). HIV risk proﬁle
of male street youth involved in survival sex. Sexually Transmitted Infections,
80(6), 526-530.

Harawa, N. T., Greenland, S., Bingham, T. A., Johnson, D. F., Cochran, S. D.,
Cunningham, W. E., et al. (2004). Associations of race/ethnicity with HIV
prevalence and HIV-related behaviors among young men who have sex with men

in 7 urban centers in the United States. Journal of Acquired Immune Deﬁciency
Syndrome, 35(5), 526-536.

Hasnain, M., Levy, J. A., Mensah, E. K., & Sinacore, J. M. (2007). Association of
educational attainment with HIV risk in African American active injection drug
users. AIDS care, 19(1), 87-91.

Hays, R. B., Kegeles, S. M., & Coates, T. J. (1990). High HIV risk-taking among young
gay men. AIDS, 4(9), 901-907.

Hays, R. 3., Paul, J ., Ekstrand, M., Kegeles, S. M., Stall, R., & Coates, T. J. (1997).
Actual versus perceived HIV status, sexual behaviors and predictors of
unprotected sex among young gay and bisexual men who identify as HIV-
negative, HIV-positive and untested. AIDS, 1 l (12), 1495-1502.

Hays, R. B., Rebchook, G. M., & Kegeles, S. M. (2003). The Mpowerment Project:
community-building with young gay and bisexual men to prevent HIV]. Am
Journal of Community Psychology, 31(3-4), 301-312.

Hein, K., Dell, R., Futterman, D., Rotheram-Borus, M. J ., & Shaffer, N. (1995).
Comparison of HIV+ and HIV- adolescents: risk factors and psychosocial
determinants. Pediatrics, 95 ( 1 ), 96-104.

Herbst, J.H., Sherba, R. T., Crepaz, N., DeLuca, J .B., Zohrabyan, L., Stall, R.D., et al.
(2005). A Meta-Analytic Review of HIV Behavioral Interventions for Reducing
Sexual Risk Behavior of Men Who Have Sex With Men. JAIDS Journal of
Acquired Immune Deficiency Syndromes, 39(2), 228-241.

Herek, G.M., & Glunt, E.K. (1988). An epidemic of stigma: Public reactions to AIDS.
American Psychologist, 43(11), 886-891.

78

 

Hirshﬁeld, S., Remien, R. H., Humberstone, M., Walavalkar, I., & Chiasson, M. A.
(2004). Substance use and high-risk sex among men who have sex with men: A
national online study in the USA. AIDS Care, 16(8), 1036-1047.

Huba, G. J., Melchior, L. A., Panter, A. T., Trevithick, L., Woods, E. R., Wright, E., et al.
(2000). Risk factors and characteristics of youth living with, or at high risk for,
HIV. AIDS Education and Prevention, 12(6), 557-575.

Janssen, R. S., Holtgrave, D. R., Valdiserri, R. 0., Shepherd, M., Gayle, H. D., & De
Cock, K. M. (2001). The serostatus approach to ﬁghting the HIV epidemic:

Prevention strategies for infected individuals. American Journal of Public Health,
91(7), 1019-1024.

Jost, S., Bernard, M. C., Kaiser, L., Yerly, S., Hirschel, B., Samri, A., et a1. (2002). A
Patient with HIV-1 Superinfection. New England Journal of Medicine, 34 7(1 0),
731.

Kahle, E. M., Freedman, M. S., & Buskin, S. E. (2005). HIV risks and testing behavior
among Asians and Paciﬁc Islanders: Results of the HIV Testing Survey, 2002-
2003. Journal of the National Medical Association, 97(Suppl), 138-188.

Kalichman, S. C., Kelly, J ., & Rompa, D. (1997). Continued high-risk sex among HIV
seropositive gay and bisexual men seeking HIV prevention services. Health
Psychology, 16(4), 369-373.

Kalichman, S. C., Kelly, J. A., & St. Lawrence, J. S. (1990). Factors inﬂuencing
reduction of sexual risk behaviors for human immunodeﬁciency virus infection: A
review. Annals of Sex Research, 3(2), 129-148.

Kalichman, S. C., Weinhardt, L., DiFonzo, K., Austin, J ., & Luke, W. (2002). Sensation
seeking and alcohol use as markers of sexual transmission risk behavior in HIV-
positive men. Annals of Behavioral Medicine, 24(3), 229-235.

Karon, J. M., Fleming, P. L., Steketee, R. W., & De Cock, K. M. (2001). HIV in the
United States at the turn of the century: an epidemic in transition. American
Journal of Public Health, 91 (7), 1060-1068.

Kass, N. E., Munoz, A., Chen, 8., Zucconi, S. L., & Bing, E. G. (1994). Changes in
employment, insurance, and income in relation to HIV status and disease

progression. The Multicenter AIDS Cohort Study. J Acquir Immune Deﬁc Syndr,
7(1), 86-91.

Kegeles, S. M., Hays, R. B., Pollack, L. M., & Coates, T. J. (1999). Mobilizing young

gay and bisexual men for HIV prevention: A two-community study. AIDS,
13(13), 1753-1762.

79

Kelly, J. A., Sikkema, K. J ., Solomon, L. J ., Heckman, T. G., Stevenson, L. Y., Norman,
A. D., et al. (1995). Factors predicting continued high-risk behavior among gay
men in small cities: psychological, behavioral, and demographic characteristics

related to unsafe sex. Journal of consulting and clinical psychology, 63(1), 101-
107.

Kelly, J. A., Sikkema, K. J ., Winett, R. A., Solomon, L. J ., & et al. (1995). Factors
predicting continued high-risk behavior among gay men in small cities:

Psychological, behavioral, and demographic characteristics related to unsafe sex.
Journal of Consulting and Clinical Psychology, 63(1), 101-107.

.Keppel, G., & Wickens, T. (2004). Design and Analysis: A Researcher ’s Handbook (4th
ed.). Upper Saddle River: Pearson Prentice Hall.

Klitzman, R. (2007). It's not just what you say: Relationships of HIV dislosure and risk
reduction among MSM in the post-HAART era. AIDS Care, 19(6), 749-756.

Kral, A. H., Lorvick, J ., Ciccarone, D., Wenger, L., Gee, L., Martinez, A., et a1. (2005).
HIV prevalence and risk behaviors among men who have sex with men and inject

drugs in San Francisco. Journal of urban health : Bulletin of the New York
Academy of Medicine, 82(1), 43-50.

Kral, A. H., Molnar, B. E., Booth, R. E., & Watters, J. K. (1997). Prevalence of sexual
risk behaviour and substance use among runaway and homeless adolescents in

San Francisco, Denver and New York City. International Journal of S TD &
AIDS, 8(2), 109-117.

Kreipe, R. E. (1985). Normal adolescent development: Helping teenagers cope with
change. New York State Journal of Medicine, 5, 214-217.

LeBlanc, A. J. (1993). Examining HIV-related Knowledge Among Adults in the US.
Journal of Health and Social Behavior, 34(1), 23-36.

Malinowsky, H. R., & Perry, G. J. (1989). AIDS information sourcebook. Phoenix,
Arizona: Oryx Press.

Malow, R. M., West, J. A., Corrigan, S. A., Pena, J. M., & Cunningham, S. C. (1994).
Outcome of psychoeducation for HIV risk reduction. AIDS Education &
Prevention, 6(2), 1 13-125.

Miller, R. L., (1995). Assisting gay men to maintain safer sex: An evaluation of an AIDS
service organization's safer sex maintenance program. AIDS Education and
Prevention, 7(Suppl), 48-63.

80

Molitor, F., Facer, M., & Ruiz, J. D. (1999). Safer sex communication and unsafe sexual
behavior among young men who have sex with men in California. Archives of
Sex Behavior, 28(4), 335-343.

Muhib, F. B., Lin, L. S., Stueve, A., Miller, R. L., Ford, W. L., Johnson, W. D., et a1.
(2001). A venue-based method for sampling hard-to-reach populations. Public
Health Reports, 116(Suppl 1), 216-222.

Myers, H. F., Javanbakht, M., Martinez, M., & Obediah, S. P. (2003). Psychosocial

predictors of risky sexual behaviors in African American men: Implications for
prevention. AIDS Education and Prevention, 15 (Suppl 1), 66-79.

Naar-King, S., Wright, K., Parsons, J ., Frey, M., Templin, T., & Ondersma, S. (2006).
Transtheoretical Model and Condom Use in HIV-Positive Youths. Health
Psychology, 25(5), 648-652.

Norton, J., Miller, R., & Johnson, M. A. (1997). Promoting HIV prevention: a problem
identiﬁcation approach to interventions in post-HIV test counselling. AIDS Care,
9(3), 345-354.

Ostrow, D.-G., McKiman, D., Klein, C., & DiFranceisco, W. (1999). Patterns and
correlates of risky behavior among HIV+ gay men: Are they really different from
HIV- men? AIDS and Behavior, 3(2), 99-110.

Paasche-Orlow, M. K., Clarke, J. G., Hebert, M. R., Ray, M. K., & Stein, M. D. (2005).
Educational attainment but not literacy is associated with HIV risk behavior
among incarcerated women. Journal of women's health, 14(9), 852-859.

Parsons, J. T., Halkitis, P. N., Wolitski, R. J ., Gomez, C. A., & the Seropositive Urban
Men's Study (2003). Correlates of sexual risk behaviors among HIV-positive men
who have sex with men. AIDS Education & Prevention, 15(5), 383-400.

Pfeifer, R. W., & Oliver, J. (1997). A study of HIV seroprevalence in a group of
homeless youth in Hollywood, California. Journal of A dolescent Health, 20(5),
339-342.

Raj, A., Cheng, D. M., Levison, R., Meli, S., & Samet, J. H. (2006). Sex Trade, Sexual
Risk, and Nondisclosure of HIV Serostatus: Findings from HIV-Infected Persons
with a History of Alcohol Problems. AIDS and Behavior, 10(2), 149-157.

Rogers, E. M. (1995). Diffusion of Innovations (4th ed.). New York, NY Free Press.
Salmon, C. T., Wooten, K., Gentry, E., Cole, G. E., & Kroger, F. (1996). AIDS
knowledge gaps: Results from the ﬁrst decade of the epidemic and implications

for future public information efforts. Journal of health communication, 1(2), 141 -
155.

81

Schiltz, M. A., & Sandfort, G. M. (2000). HIV—positive people, risk and sexual
behaviour. Social Science & Medicine, 50(11), 1571-1588.

Semple, S. J ., Patterson, T. L., & Grant, I. (2000). The sexual negotiation behavior of

HIV-positive gay and bisexual men. Journal of consulting and clinical
psychology, 68(5), 934-937.

Smith, D. M., Wong, J. K., Hightower, G. K., Ignacio, C. C., Koelsch, K. K., Daar, E. S.,
et al. (2004). Incidence of HIV Superinfection Following Primary Infection.
Journal of the American Medical Association, 292(10), 1177-1178.

Somlai, A. M., Kalichman, S. C., & Bagnall, A. (2001). HIV risk behaviour among men
who have sex with men in public sex environments: an ecological evaluation.
AIDS Care, 13(4), 503-514.

Stall, R., & Purcell, D.-W. (2000). Intertwining epidemics: A review of research on
substance use among men who have sex with men and its connection to the AIDS
epidemic. AIDS and Behavior, 4(2), 181-192.

Stein, J. A., Rotheram-Borus, M. J ., Swendeman, D., & Milbum, N. G. (2005). Predictors
of sexual transmission risk behaviors among HIV-positive young men. AIDS
Care, 17(4), 433-442.

Stine, G. J. (2005). AIDS update: An overview of acquired immune deﬁciency syndrome
(2005) San Francisco: Pearson Benjamin Cummings.

Tabachnick, B. G., & F idell, L. S. (2006). Using Multivariate Statistics. Boston: Allyn
and Bacon.

Valleroy, L. A., MacKellar, D. A., Karon, J. M., Rosen, D. H., McFarland, W., Shehan,
D. A., et al. (2000). HIV prevalence and associated risks in young men who have
sex with men. JAMA: Journal of the American Medical Association, 284(2), 198-
204.

Vitry-Henry, L., Penalba, C., Beguinot, I., & Deschamps, F. (2002). Relationships
between work and HIV/AIDS status. Occupational Medicine, 49(2), 115-116.

Weinhardt, L.S., Carey, M.P., Johnson, B.T., & Bickham, N.L. (1999). Effects of HIV
counseling and testing on sexual risk behavior: A meta-analytic review of
published research, 1985-1997. American Journal of Public Health, 89(9), 1397-
1405.

Whitbeck, L. B., Chen, X., Hoyt, D. R., Tyler, K. A., & Johnson, K. D. (2004). Mental
Disorder, Subsistence Strategies, and Victimization Among Gay, Lesbian, and

82

Bisexual Homeless and Runaway Adolescents. Journal of Sex Research, 41(4).
329-342.

Wolitski, R. J., Flores, S. A., O'Leary, A., Bimbi, D. S., & Gomez, C. A. (2007). Beliefs
about personal and partner responsibility among HIV-seropositive men who have

sex with men: measurement and association with transmission risk behavior.
AIDS & Behavior, 11(5), 676-686.

Wolitski, R. J., Henny, K. D., Lyles, C. M., Purcell, D. W., Carey, J. W., Crepaz, N., et
al. (2006). Evolution of HIV/AIDS Prevention Programs--United States, 1981-
2006. JAMA: Journal of the American Medical Association, 296(7), 760-762.

Zaidi, I. F., Crepaz, N., Song, R., Wan, C. K., Lin, L. S., Hu, D. J., et al. (2005).

Epidemiology of HIV/AIDS among Asians and Paciﬁc Islanders in the United
States. AIDS Education and Prevention, 17(5), 405-417.

83

FOOTNOTES

I Some variation from the seasonal timeframe occurred in Detroit due to
administrative problems.

2 It was originally thought that city must be controlled because HIV
seroprevalence varies greatly by the communities included in the sample. However, city
was not included because, due to the fact that that the cities were stratiﬁed by race, there
was a strong correlation between the city and race variables. Thus, including this variable
would have provided redundant information and may have introduced problems with
multi-collenearity. In addition, it was not possible to include this variable because there
were not enough HIV-positive men in some cities to achieve the sufﬁcient cell frequency
needed for statistical analyses. Finally, separate analyses revealed that community did not
signiﬁcantly predict HIV-status after race was entered into a logistic regression model.

3 The main analyses were also performed without the men reporting an
“indeterminate” HIV-status. No signiﬁcant differences in our ﬁndings were detected as a

result of this analysis.

84

   

llllllllllllllllll