AIDS Behav (2015) 19:257–269

DOI 10.1007/s10461-014-0937-2

ORIGINAL PAPER

Sexual Orientation- and Race-Based Discrimination and Sexual

HIV Risk Behavior Among Urban MSM
Victoria Frye • Vijay Nandi • James Egan • Magdalena Cerda • Emily Greene •

Hong Van Tieu • Danielle C. Ompad • Donald R. Hoover • Debbie Lucy •

Eduardo Baez • Beryl A. Koblin

Published online: 9 November 2014

Ó Springer Science+Business Media New York 2014

Abstract Understanding what social factors are associ- self-reported HIV-positive or unknown status (377), 7 %
ated with risk of HIV acquisition and transmission among (N = 27) reported having unprotected insertive anal
gay, bisexual and other men who have sex with men intercourse with an HIV-negative or unknown status part-
(MSM) is a critical public health goal. Experiencing dis- ner (‘‘HIV transmission risk’’). Of MSM who self-reported
crimination may increase risk of HIV infection among HIV-negative status (992), 11 % (110) reported unpro-
MSM. This analysis assessed relations between experi- tected receptive anal intercourse with an HIV-positive or
ences of sexual orientation- and race-based discrimination unknown status partner (‘‘HIV acquisition risk’’). HIV
and sexual HIV risk behavior among MSM in New York acquisition risk was positively associated with sexual ori-
City. 1,369 MSM completed a self-administered comput- entation-based discrimination in home or social neighbor-
erized assessment of past 3-month sexual behavior, expe- hoods, but not race-based discrimination. We observed that
rience of social discrimination and other covariates. sexual orientation-based discrimination was associated
Regression models assessed relations between recent with sexual HIV risk behavior among urban-dwelling
experience of discrimination and sexual HIV risk behavior. MSM. Addressing environmental sources of this form of
Mean age was 32 years; 32 % were white; 32 % Latino/ discrimination, as well as the psychological distress that
Hispanic; 25 % African American/Black. Of MSM who may result, should be prioritized in HIV prevention efforts.

Keywords HIV prevention
Social discrimination

Sexual behavior
MSM

For the NYCM2M Study Team.

V. Frye (&)
E. Greene M. Cerda

Laboratory of Social and Behavioral Sciences, Lindsley F. Department of Epidemiology, Mailman School of Public Health,
Kimball Research Institute, New York Blood Center, 310 East Columbia University, New York, NY, USA
67th Street, New York, NY 10065, USA
e-mail: vfrye@nybloodcenter.org H. Van Tieu
D. Lucy
E. Baez
B. A. Koblin
Laboratory of Infectious Disease Prevention, Lindsley F.
V. Frye Kimball Research Institute, New York Blood Center, New York,
Department of Sociomedical Sciences, Mailman School of NY, USA
Public Health, Columbia University, New York, NY, USA
D. C. Ompad
V. Nandi Department of Nutrition, Food Studies, and Public Health,
Data Management Services, Lindsley F. Kimball Research Steinhardt School of Culture, Education and Human
Institute, New York Blood Center, New York, NY, USA Development, New York University, New York, USA

J. Egan
E. Greene D. R. Hoover

Department of Behavioral and Community Health Sciences, Department of Statistics and Biostatistics and Institute for
Graduate School of Public Health, University of Pittsburgh, Health, Health Care Policy and Aging Research, Rutgers, The
Pittsburgh, USA State University of New Jersey, Piscataway, NJ, USA

123
258 AIDS Behav (2015) 19:257–269

Introduction race system operates to maintain the marginalized status of

racial and ethnic minorities [21]. A robust body of litera-
In the United States (US), gay, bisexual and other men who ture documents the positive relationship between experi-
have sex with other men (MSM) continue to be dispro- ences of racial discrimination and poor health (for an
portionately affected by HIV, accounting for 61 % of new overview see Williams and Mohammad 2009) [22].
cases in 2010 [1, 2]. Young Black MSM were the only Social stress theory helps explain how social discrimi-
group in the US with a statistically significant increase nation experience translates into adverse health behaviors
(48 %) in new HIV infections between 2006 and 2009 [2]. and ultimately poor health outcomes. Sexual minority
In New York City (NYC) during 2011, 51 % of new HIV stress theory posits that certain environments both increase
diagnoses were among MSM, and new diagnoses among the likelihood of experiencing personal stress events (i.e.,
MSM younger than 30 increased from 2001 to 2010. Fur- traumatic events, daily hassles) and provide resources to
ther, young African American or Black MSM accounted cope with such stressors [23, 24]. Extending social stress
for approximately 50 % of new HIV diagnoses among theory [25, 26], Meyer posits that sexual minorities expe-
young MSM living in NYC [3]. Despite limited evidence rience chronic stress in their daily interactions due to being
of the importance of the roles of individual-level risk members of stigmatized groups, including social discrim-
behavior, for example number of sexual partners and drug ination based on perceived sexual orientation or gender
use [4–6], in explaining racial disparities in HIV incidence identity [27]. Increased stress due to experience of sexual
and prevalence, the role of sexual behavior in HIV trans- orientation-based discrimination may contribute to
mission is crucial to epidemic propagation among MSM depression, drug and alcohol use and/or sexual compul-
[7]. Identifying personal, social and environmental factors sivity as a coping mechanism [28]. The negative health
associated with the sexual behavior that increases HIV effects of sexuality-based discrimination include psycho-
acquisition and transmission risk among MSM is critical to logical distress [29], drug and alcohol use [30], and risky
informing prevention policy and practice. sexual behavior [31, 32]. Associations between sexual
Increasingly, social and environmental factors are being orientation-based discrimination and sexual risk behavior
assessed as contributors to sexual behavior that increases have been found in several recent studies among MSM [31,
risk of HIV infection [8], among MSM specifically [9–12]. 33–35].
A recent meta-analysis ranked correlates of HIV infection Gay, bisexual and other MSM who are also racial or
among MSM and found that sociostructural factors con- ethnic minorities potentially experience the ‘‘double
stituted the majority of the top 10 correlates, including low whammy’’ of both the sex-gender-sexuality and the race
income and education, recent unemployment, and history systems [15] and associated racial discrimination and
of incarceration, with African American or Black MSM negative life events, leading to increased risk of adverse
being twice as likely to report experiencing these factors health outcomes [36–38]. Individuals who are affected by
than other MSM [13]. These sociostructural factors repre- both race and sexual orientation systems [39] are said to
sent static markers of social systems [14, 15] that mar- possess ‘‘dual minority status’’ and thus may be at partic-
ginalize specific subpopulations and produce adverse ularly high risk of adverse health outcomes, such as HIV
health outcomes at the population level. Over the life- infection. Recent reports on the relationship between both
course, among MSM, intersecting systems of race and forms of social discrimination and sexual risk behavior
sexual orientation increase the chances of exposures and assess how sexual orientation- and race-based discrimina-
experiences that result in negative health outcomes, termed tion work in conjunction. Mizuno and colleagues [31]
syndemics [16, 17]. found that Latino MSM who experienced both sexual ori-
entation- and race-based discrimination were more likely to
Sexual Orientation- and Race-Based Discrimination report unprotected anal intercourse (with a casual sex
partner) and binge alcohol drinking, as compared to Latino
Sexual orientation-based discrimination manifests the het- men who did not experience either form of discrimination.
eronormativity and homophobia that characterize the val- Diaz et al. (2004) found among Latino MSM that risky
ues and attitudes inherent in the sex-gender-sexuality sexual situations mediated the relationship between both
system [18]. This system depends upon the cultural stigma social oppression and psychological distress and high HIV
associated with homosexuality and gender identity devi- risk [40]. Another analysis among both Latino and Black
ance to legitimize its oppressive function [19, 20]. Men MSM found statistically significant paths between experi-
who identify as gay or bisexual, engage in same-sex sexual ences of both sexual orientation- and race-based discrimi-
behavior, and/or are gender non-conforming, transgress the nation and sero-discordant unprotected anal intercourse
prescribed heterosexual and masculine norms; thus, they (SUAI) mediated by ‘‘risky sexual situations’’ (e.g., sex
are often targets for acts of discrimination. Similarly, the while high or survival sex) among Black MSM. However,

123
AIDS Behav (2015) 19:257–269 259

among Latino MSM, significant paths were found only enclaves, those with a growing gay population, and
between experiences of sexual orientation-based discrimi- neighborhoods with a much less visible or documented gay
nation and SUAI [33]. presence. The internet- and mobile application-based
This analysis assessed the relationship between sexual recruitment strategy was added in July 2012 in response to
orientation- and race-based experiences of discrimination the proliferation of apps that appeared since recruitment for
and sexual HIV risk behavior among a large and diverse the study began. Banner and pop-up ads were placed on
sample of men living in a major urban area. Like previous various websites and apps every 2–3 months until the study
analysts, we examined race- and sexual orientation-based was fully enrolled. In the sample presented here, 56 %
social discrimination separately and in combination, how- (724) of MSM were recruited using face-to-face methods
ever we extend prior work by exploring whether place of and 44 % (645) were recruited via the internet or using
experienced discrimination, specifically participants’ self- mobile apps.
defined home and social neighborhoods, relates to sexual Individuals were eligible to participate if they reported
HIV risk behavior. Additionally, we stratified our analyses being a biological male at birth, were at least 18 years of
by self-reported HIV status and perceived partner HIV age, resided in NYC, reported engaging in anal sex with a
status to examine potential acquisition and transmission man in the past 3 months, communicated in English or
risks, as the literature has shown that HIV-positive MSM Spanish and were willing and able to give informed con-
reduce their sexual risk behavior significantly once they sent for the study. Thus, 4,998 men were approached and
learn their status to avoid transmission to their partners provided contact information; 1,997 men met the study’s
[41]. Further, we assessed factors that may explain the eligibility criteria and scheduled a study visit and 1,503
connection between experiences of discrimination and men enrolled (75 %). After excluding 21 men who did not
sexual HIV risk behavior, such as internalized homopho- report any sex partners in the past 3 months and 107 men
bia, psychological distress and sex while under the influ- with significant missing data, 1,369 MSM were included in
ence of alcohol and/or drugs. In addition, we controlled for the present analysis. Institutional review boards at the New
a range of psychosocial factors that have been found to be York Blood Center, New York Academy of Medicine and
independently associated with sexual HIV risk behavior. New York University reviewed the study. After providing
These included identity-related factors, such as race/eth- informed consent, participants met with a staff member to
nicity-related identity factors, which consistent with social complete the Neighborhood Locator Questionnaire which
stress theory may buffer the negative effects of social collected information on the location of their home (where
discrimination on sexual risk behavior [42, 43], and sexu- they live) and social (where they socialize most often)
ality-related identity and attachment factors, which may act neighborhoods. All other data collected, other than HIV
to increase risk behavior [44–46]. We also controlled for testing, were gathered by participant self-report using audio
peer norms and condom use self-efficacy, two factors that computer-assisted self-interview (ACASI) technology.
have consistently been found to be highly correlated with Upon completion of the visit, participants received $50 and
sexual HIV risk behavior among MSM [47, 48]. Finally, a two-way Metrocard for their time and transportation
our analysis was limited to African American or Black and costs.
Latino MSM in order to examine unique correlates of
sexual HIV risk behavior for minority men. Measures

Primary Dependent Measures

Methods
Sexual behavior data collected included total number of
Sampling and Study Eligibility male, female and transgender partners and total number of
anal sex acts by partner HIV status, number of acts where
Details of the NYC M2M study procedures have been condoms were used and use of drugs/alcohol in conjunction
published previously [49]. Briefly, from October 2010 to with sex in the three months prior to study. In this analysis,
July 2013, MSM in New York City were recruited using a we created two outcome variables based on self-reported
modified venue-based time–space sampling methodology sexual behavior and self-reported HIV status: HIV acqui-
for both physical (e.g., street locations, bars/clubs, and sition risk behavior and HIV transmission risk behavior.
retail businesses, etc.) and virtual (e.g., geospatial net- Among 992 HIV-negative participants, HIV acquisition
working applications, etc.) venues [50]. Locations and risk behavior was coded as 1 among HIV-negative partic-
associated day-time periods were randomly selected each ipants who had unprotected receptive anal sex with any
month from a sampling frame that included a wide range of type (e.g., primary, casual, etc.) of HIV-positive or
neighborhoods that are traditionally considered gay unknown HIV status male sex partner; the remaining

123
260 AIDS Behav (2015) 19:257–269

HIV-negative participants were coded as 0. Among the sex. Psychological distress was operationalized as a com-
remaining 377 participants, HIV transmission risk behavior bination of measures of depression and anxiety, using
was coded as 1 among HIV-positive or unknown status adapted versions of the Patient Health Questionnaire
participants who had unprotected, insertive anal sex with (PHQ-9) and the Generalized Anxiety Disorder Assess-
any HIV-negative or unknown status male sex partner; the ment (GAD-7) [51, 52], respectively. The PHQ-9 is a brief
remaining HIV-positive or unknown status participants depression scale, which we adapted to ask participants
were coded as 0. whether they experienced one of nine symptoms for a
2-week period during the past 3 months (yes/no). Simi-
Primary Independent Measures larly, the GAD-7 is a brief measure of anxiety disorder,
which we adapted to ask participants whether they expe-
Experience of race- and sexual orientation-based discrim- rienced one of the seven symptoms for a 2-week period
ination in the participant’s home and social neighborhoods during the past 3 months (yes/no). For use in this analysis,
were assessed in the 3 months prior to interview. Racial the total number of symptoms experienced for a 2-week
discrimination was assessed using the following question: period in the past 3 months was summed. Internalized
‘‘Have you experienced discrimination, been prevented homophobia was assessed using Herek’s scale (1984) using
from doing something or been hassled or been made to feel 5-point Likert scale with answer choices ranging from
inferior because of your race, ethnicity or color?’’. Sexual strongly disagree to strongly agree and was modeled as a
orientation-based discrimination was assessed using the continuous variable [53, 54]. The internal consistency was
following question: ‘‘Have you experienced discrimination, 0.89. Sex while drunk or high/having used drugs was
been prevented from doing something or been hassled or assessed with two questions regarding last anal sex, ‘‘The
been made to feel inferior because of your sexual orien- last time you had anal sex, were you buzzed or drunk on
tation?’’. These questions were asked in reference to both alcohol?’’ and ‘‘The last time you had anal sex, did you use
the home (where they lived) and social (where they spent any drugs within two hours before or during the time you
most of their time socializing) neighborhoods of the par- had sex?’’ If a respondent responded yes to either question
ticipants. Using these questions we created a 4-category the value assigned was 1; if they responded no to both the
independent variable reflecting: (1) experience of neither value assigned was 0.
sexual orientation- nor race-based discrimination; (2) Measures of identity and affiliation included racial/eth-
experience of only sexual orientation-based discrimination; nic identity using the Multigroup Ethnic Identity Measure
(3) experience of only race-based discrimination; or (4) (MEIM) [55] and gay community attachment [42]. The
experience of both sexual orientation- and race-based dis- MEIM is a 21-item scale that uses a 4-point Likert scale
crimination. We created variables for each neighborhood with responses ranging from strongly disagree to strongly
type separately and in combination, that is, the four-cate- agree. The measure of gay community attachment was
gory discrimination outcome in either the home or social assessed using a 12-item scale using a 4-point Likert scale
neighborhood. In multivariable analyses, experience of with responses ranging from strongly disagree to strongly
neither form of social discrimination was the referent agree. Both were modeled as continuous variables and
category. internal consistencies were Cronbach’s alpha = 0.85 and
0.81, respectively. Degree of ‘‘outness’’ about sexual
Sociodemographic Measures identity to friends and family was measured with a single
item assessing how ‘‘out’’ participants were on a scale of
Demographic characteristics assessed included measures 1–10, with 1 being ‘‘not out to anyone’’ and 10 being ‘‘out
for age, race/ethnicity, education, employment status, to everyone.’’ Finally, condom use self-efficacy [56] and
annual personal income, income insufficiency for basic perceived peer sexual risk norms [57] were also assessed,
needs (i.e., food, shelter, and utilities) and partnership using 11- and 7-item scales, respectively, and 5- and
status (i.e., married or registered domestic partnership). All 4-level Likert responses, because of their strong association
sociodemographic measures were modeled as categorical with condom use behavior among MSM [58–60]; both
variables. were modeled as continuous variables. Internal consisten-
cies were acceptable, with 0.87 and 0.84 for condom use
Psychosocial and Condom Use-Related Measures self-efficacy and perceived peer norms, respectively.

Factors that may link experiences of sexual orientation- Analysis

based discrimination to sexual risk behavior include psy-
chological distress, internalized homophobia and sex while Data from this cross-sectional study were analyzed in SAS
buzzed on alcohol or having used drugs before or during software (version 9.3) for Windows (SAS Institute Inc.,

123
AIDS Behav (2015) 19:257–269 261

Cary NC, USA). Unadjusted associations between the The mean number of sex partners in the past 3 months was
primary independent variables, sociodemographic factors, 5.2 (SD = 7.8) and the median was 3 (IQR = 2, 5). Of the
psychosocial factors (e.g., psychological distress, internal- HIV-positive or unknown status men (377), 7 % (N = 27)
ized homophobia, AOD use before sex, peer norms, etc.) reported unprotected insertive anal intercourse with a
and the primary dependent variable were assessed using v2, partner who was HIV-negative or unknown status (‘‘HIV
t tests, one-way ANOVAs, and Mann–Whitney non-para- transmission risk’’). We did not find any statistically sig-
metric tests, as appropriate. The bivariate significance level nificant differences in the primary outcomes by recruitment
was set at p \ .05 for inclusion in the multivariable model. method (venue- vs. Web-based) (data not shown). Of the
Continuous measures were standardized so that the odds HIV-negative MSM (992), 11 % (110) reported unpro-
ratios reflect one standard deviation change in the score of tected receptive anal intercourse with an HIV-positive or
the measure. In building our multivariable logistic regres- unknown status partner (‘‘HIV acquisition risk’’). Over a
sion models, we began by estimating the crude association third (36 %) of men reported using drugs or alcohol at last
between the primary independent variables and the out- sex. Mean internalized homophobia was 1.7 (SD = 0.8),
comes. Next, we added psychosocial variables in concep- approaching the ‘‘disagree’’ response; mean psychological
tual sets, starting with the factors that might link our distress was 4.6 (SD = 4.6) out of a possible of 16.
primary independent and dependent variable (i.e., psy- In the past 3 months, 15 % of men reported experienc-
chological distress, alcohol and/or drug use before or dur- ing either sexual orientation- or race/ethnicity-based dis-
ing sex and internalized homophobia). Next we added crimination in their home or social neighborhood; 5 %
sociodemographic factors to the model; finally we added reported sexual orientation-based discrimination only and
the remaining psychosocial and condom use-related fac- another 5 % reported race-based discrimination only. Six
tors. Factors that did not retain statistical significance in the percent reported experiencing both forms of discrimination
model were removed after each step. in their home or social neighborhoods. Upon examination
of distribution of discrimination experiences by home and
social neighborhood, fewer participants reported experi-
Results encing either form of discrimination in their social neigh-
borhoods, as compared with their home neighborhoods. We
Univariate Results did not observe differences in the direction of the estimates
of association among social discrimination and the out-
The average age was 32.0 (SD = 10.3); 32 % of the comes by neighborhood (home vs. Social) where the dis-
sample were white (non-Hispanic); 32 % Hispanic; 25 % crimination took place (Table 2).
Black/African American and 13 % reported another eth-
nicity, such as Asian, Native American Indian, etc. Nearly Bivariate Results
half (49 %) of men reported having a college degree or
more and another third had some college education; just Experience of sexual orientation-based discrimination
6 % had less than a high school degree. The plurality of only, but not race-based or both forms of discrimination, in
men (40 %) worked full-time and 24 % worked part-time; either the home or social neighborhood, was significantly
less than a third (30 %) was not working. Just over a associated with sexual HIV acquisition risk behavior
quarter (26 %) reported an average personal income of less (UOR = 3.36; 95 % CI 1.71, 6.61). Experiencing only
than $10,000 per year; 42 % reported an income of race-based discrimination or both race- and sexual orien-
$10,000–39,999 and 32 % reported an income of $40,000 tation-based discrimination was not significantly associated
or greater. Financial insecurity affected nearly half of the with acquisition risk behavior. Estimates of association
sample with 48 % reporting that they sometimes did not with transmission risk behavior were unstable due to small
have enough money for rent, food utilities and other basic numbers and thus are not presented (Table 2).
needs. Only 4 % of men sampled reported being married or Of the factors we considered to potentially link experi-
in a registered domestic partnership with another man. The ence of social discrimination and sexual risk behavior,
majority of men (88 %) self-identified as exclusively gay alcohol or drug use before or during last sex (uOR = 2.01;
or homosexual; 9 % self-identified as bisexual and 3 % 95 % CI 1.35, 2.99), psychological distress (uOR = 1.65;
identified as straight/heterosexual or ‘‘other’’. Average 95 % CI 1.37, 1.98), and internalized homophobia
‘‘outness’’ was 8.2 (range: 1–10; SD = 3.3), indicating that (uOR = 1.22; 95 % CI 1.01, 1.46) were significantly
most men were ‘‘out’’ to most people they know (Table 1). associated with acquisition risk in bivariate analyses. Just
Seventy-three percent (992) of the sample reported that one of the sociodemographic factors, financial insecurity,
they were HIV-negative; 23 % (312) reported being HIV- and two of the psychosocial factors that we considered as
positive and 5 % (65) did not know or refused to answer. potential independent correlates of unprotected sex,

123
262 AIDS Behav (2015) 19:257–269

Table 1 Sociodemographic, sexual behavior, and identity- and attachment-related characteristics and HIV acquisitiona and transmissionb risk
behavior, M2MNYC, N = 1,369
Characteristics Totalc Acquisition risk p value Transmission risk p value
(N = 992) (N = 377)
Yes No Yes No
N (%) N (%) N (%) N (%) N (%)

Recruitment NS NS
Venue 724 (52.9) 65 (59.1) 459 (52.0) 12 (44.4) 188 (53.7)
Online 645 (47.1) 45 (40.9) 423 (48.0) 15 (55.6) 162 (46.3)
Sociodemographics
Age (Mean, SD) 32.0 (10.3) 31.5 (11.0) 30.6 (9.6) NS 39.6 (11.0) 34.9 (10.9) 0.033
Age, categorized NS NS
18–24 358 (26.1) 30 (27.3) 257 (29.2) 2 (7.4) 69 (19.7)
25–29 368 (26.9) 37 (33.6) 251 (28.5) 3 (11.1) 77 (22.0)
30–39 330 (24.1) 21 (19.1) 222 (25.2) 8 (29.6) 79 (22.6)
40? 312 (22.8) 22 (20.0) 151 (17.1) 14 (51.9) 125 (35.7)
Race/ethnicity NS 0.001
White 431 (31.6) 39 (35.5) 319 (36.3) 13 (48.2) 60 (17.2)
Black 334 (24.5) 22 (20.0) 186 (21.2) 6 (22.2) 120 (34.4)
Hispanic 429 (31.5) 31 (28.2) 259 (29.5) 7 (25.9) 132 (37.8)
All other 170 (12.5) 18 (16.4) 114 (13.0) 1 (3.7) 37 (10.6)
Education NS NS
Less than high school graduate 78 (5.7) 6 (5.5) 29 (3.3) 1 (3.7) 42 (12.0)
High school graduate 152 (11.1) 16 (14.5) 83 (9.4) 3 (11.1) 50 (14.3)
Some college 468 (34.2) 29 (26.4) 284 (32.2) 11 (40.7) 144 (41.1)
College graduate or more 671 (49.0) 59 (53.6) 486 (55.1) 12 (44.4) 114 (32.6)
Employment 0.075 NS
Working full-time 547 (40.0) 48 (43.6) 405 (46.0) 10 (37.0) 84 (24.1)
Working part-time 322 (23.6) 23 (20.9) 221 (25.1) 6 (22.2) 72 (20.6)
Not working, looking/not working, 413 (30.2) 37 (33.6) 209 (23.8) 8 (29.6) 159 (45.6)
not looking/temporarily laid off/retired
Working off the book/other 84 (6.1) 2 (1.8) 45 (5.1) 3 (11.1) 34 (9.7)
Personal Income NS 0.020
\$10,000 348 (25.9) 32 (29.1) 194 (22.4) 7 (25.9) 115 (33.5)
$10,000–39,999 561 (41.7) 40 (36.4) 348 (40.2) 9 (33.3) 164 (47.8)
$40,000–59,999 209 (15.5) 14 (12.7) 163 (18.8) 3 (11.1) 29 (8.5)
$60,000? 228 (16.9) 24 (21.8) 161 (18.6) 8 (29.6) 35 (10.2)
Financial insecurity
Not enough $ for rent, food, or utilities 649 (47.7) 59 (53.6) 378 (43.1) 0.035 16 (59.3) 196 (56.7) NS
Not enough $ for social activity 951 (69.6) 81 (73.6) 590 (67.0) NS 18 (66.7) 262 (75.3) NS
Partnership status
Married or registered domestic partner 60 (4.4) 5 (12.8) 34 (87.2) NS 0 (0.0) 21 (6.0) NS
Sexual Identity NS NS
Gay, homosexual, queer, same gender 1,201 (87.7) 99 (90.0) 764 (86.6) 26 (96.3) 312 (89.1)
loving, etc.
Bisexual 125 (9.1) 9 (8.2) 93 (10.5) 1 (3.7) 22 (6.3)
Straight, heterosexual 8 (0.6) 0 (0.0) 6 (0.7) 0 (0.0) 2 (0.6)
Unsure or questioning/other/missing 35 (2.6) 2 (1.8) 19 (2.2) 0 (0.0) 14 (4.0)
Psychosocial Factors
‘‘Outness’’ (Mean, SD) 8.2 (3.3) 8.2 (2.3) 8.2 (2.2) NS 8.8 (2.0) 8.3 (5.3) NS
Racial/ethnic identity (Mean, SD) 3.0 (0.6) 2.9 (0.6) 3.0 (0.6) NS 3.1 (0.6) 3.1 (0.5) NS

123
AIDS Behav (2015) 19:257–269 263

Table 1 continued
Characteristics Totalc Acquisition risk p value Transmission risk p value
(N = 992) (N = 377)
Yes No Yes No
N (%) N (%) N (%) N (%) N (%)

Gay community attachment (Mean, SD) 3.2 (0.5) 3.1 (0.5) 3.2 (0.5) NS 3.3 (0.4) 3.2 (0.5) NS
Condom use-related factors
Condom use self-efficacy (Mean, SD) 4.2 (0.6) 3.8 (0.6) 4.3 (0.6) \.001 3.8 (0.4) 4.1 (0.7) 0.009
Perceived peer condom use norms (Mean, SD) 2.9 (0.6) 2.8 (0.5) 2.9 (0.5) 0.012 2.5 (0.7) 2.9 (0.6) 0.002
a
HIV acquisition risk behavior was coded as 1 among HIV-negative participants who had unprotected receptive anal sex with any type (e.g.,
primary, casual, etc.) of HIV-positive or unknown HIV status male sex partner; the remaining HIV-negative participants were coded as 0
b
HIV transmission risk behavior was coded as 1 among HIV-positive or unknown status participants who had unprotected, insertive anal sex
with any HIV-negative or unknown status male sex partner; the remaining HIV-positive or unknown status participants were coded as 0
c
N do not total to 1,369 due to missing data

Table 2 Unadjusted associations among sexual orientation- and were either not significantly associated or estimates were
race-based discrimination (past 3 months) and HIV acquisitiona and unstable, due to low numbers (Table 3).
transmissionb risk behavior, M2MNYC, N = 1,369
Characteristics Total N (%) Acquisition risk p value Multivariate Results
(N = 992)
OR 95 % CI The final model included the 4-category primary indepen-
dent variable, with experience of sexual orientation-based
Home neighborhood only
discrimination only in either the home or social neighbor-
None 1,204 (88.0) ref 0.038
hood being significantly associated with sexual HIV
Sexual orientation- 57 (4.2) 2.87 1.36, 6.07
based only
acquisition risk behavior (aOR = 2.50; 95 % CI 1.17,
Race-based only 56 (4.0) 1.48 0.61, 3.61
5.35). Psychological distress (aOR = 1.43; 95 % CI 1.17,
1.76), alcohol or drug use before or during last sex
Both 52 (3.8) 0.78 0.24, 2.61
(aOR = 1.76; 95 % CI 1.13, 2.72) and condom use self-
Social neighborhood only
efficacy (aOR = 0.47; 95 % CI 0.38, 0.58) were all inde-
None 1,270 (92.8) ref 0.049
pendently and significantly associated with the outcome
Sexual orientation- 28 (2.1) 3.80 1.41, 10.23
based only (Table 4).
Race-based only 43 (2.5) 1.03 0.31, 3.48 Finally, we ran models on the sample including only
Both 36 (2.6) 0.41 0.06, 3.10 African American/Black or Latino participants (N = 498).
Either home or social neighborhood
Bivariate analyses revealed that unlike in the full sample,
None 1,163 (85.0) ref 0.002
financial insecurity and perceived peer norms around
condom use were not associated with acquisition risk
Sexual orientation- 63 (4.6) 3.36 1.71, 6.61
based only among African American and Latino participants
Race-based only 66 (4.8) 1.67 0.76, 3.68 (Table 5). As with the full sample, we found a pattern of
Both 67 (5.6) 0.57 0.17, 1.88 association where the odds of acquisition risk increased
with the experience of sexual orientation-based discrimi-
Odds ratios reflect change of 1 standard deviation
a
nation (uOR = 2.24; 95 % CI 0.86, 5.80), but the associ-
HIV acquisition risk behavior was coded as 1 among HIV-negative
ation was not statistically significant (Table 6).
participants who had unprotected receptive anal sex with any type (e.g.,
primary, casual, etc.) of HIV-positive or unknown HIV status male sex
partner; the remaining HIV-negative participants were coded as 0
b
HIV transmission risk behavior was coded as 1 among HIV-positive Discussion
or unknown status participants who had unprotected, insertive anal sex
with any HIV-negative or unknown status male sex partner; the
In this racially and ethnically diverse sample, we found
remaining HIV-positive or unknown status participants were coded as 0
that 15 % of participants reported experiencing either race-
and/or sexual orientation-based discrimination in either
condom use self-efficacy and perceived peer sexual risk their home or social neighborhoods. Participants reported
norms, were statistically significantly related to acquisition similar levels of the two forms of discrimination, yet it was
risk behavior. In terms of transmission risk, the factors self-reported experience of sexual orientation-based

123
264 AIDS Behav (2015) 19:257–269

Table 3 Unadjusted associations among select psychosocial factors and HIV acquisitiona and transmissionb risk behavior, M2MNYC,
N = 1,369
Total N (%) Acquisition risk (N = 992) p value Transmission risk (N = 377) p value
OR 95 % CI OR 95 % CI

Alcohol and/or drug use before/during sex 490 (35.8) 2.01 1.35, 2.99 \0.001 0.80 0.36, 1.80 NS
Psychological distress (Mean, SD)1 4.6 (4.6) 1.65 1.37, 1.98 \0.001 1.26 0.87, 1.83 NS
Internalized homophobia (Mean, SD)1 1.7 (0.8) 1.22 1.01, 1.46 0.038 0.64 0.38, 1.07 NS
a
HIV acquisition risk behavior was coded as 1 among HIV-negative participants who had unprotected receptive anal sex with any type (e.g.,
primary, casual, etc.) of HIV-positive or unknown HIV status male sex partner; the remaining HIV-negative participants were coded as 0
b
HIV transmission risk behavior was coded as 1 among HIV-positive or unknown status participants who had unprotected, insertive anal sex
with any HIV-negative or unknown status male sex partner; the remaining HIV-positive or unknown status participants were coded as 0
1
Odds ratios reflect change of 1 standard deviation

Table 4 Adjusted associations among social discrimination and HIV acquisitiona risk behavior, M2MNYC, N = 937
MODEL #1 MODEL #2 MODEL #3
AOR (95 % CI) AOR (95 % CI) AOR (95 % CI)

Discrimination (P3M)
None Reference Reference Reference
Sexual orientation-based only 3.36 (1.71, 6.61) 2.54 (1.26, 5.12) 2.50 (1.17, 5.35)
Race/ethnicity-based only 1.67 (0.76, 3.68) 1.39 (0.61, 3.14) 1.27 (0.53, 3.06)
Both sexual orientation- and race/ethnicity-based 0.57 (0.17, 1.88) 0.47 (0.14, 1.56) 0.43 (0.12, 1.60)
Psychosocial factors
Psychological distress1 1.61 (1.33, 1.94) 1.43 (1.17, 1.76)
Alcohol and/or drug use before/during sex 1.77 (1.17, 2.67) 1.76 (1.13, 2.72)
Condom use-related factors
Condom use self efficacy1 0.47 (0.38, 0.58)
a
HIV acquisition risk behavior was coded as 1 among HIV-negative participants who had unprotected receptive anal sex with any type (e.g.,
primary, casual, etc.) of HIV-positive or unknown HIV status male sex partner; the remaining HIV-negative participants were coded as 0
1
Odds ratios reflect change of 1 standard deviation

discrimination only within the past 3 months that was ratio of sexual orientation-based discrimination in Model 2
significantly associated with sexual HIV acquisition risk as compared with Model (Table 4). Empirical evidence is
behavior, controlling for known psychosocial correlates. accumulating that psychological distress and substance
This result is consistent with prior work among MSM [29– abuse are outcomes of experiences of sexual orientation-
31, 33]. When we restricted the sample to African Amer- based discrimination among LGBTQ individuals [62–65].
ican and Latino men, we found that sexual orientation- In our sample, these two factors may partially explain the
based discrimination only was associated with sexual HIV association between social discrimination and sexual HIV
acquisition risk behavior, but it was not statistically sig- risk behavior among MSM, which is generally consistent
nificant. This result may have been due to the smaller with sexual minority stress theory [27]. However, further
sample size and reduced power to detect the association longitudinal research is needed to assess causal relations
among the restricted sample; other studies where such an among these factors.
association was found have had larger samples [31, 33, We did not find support for the role of internalized
61]. homophobia as a correlate of HIV acquisition risk behav-
Factors that may link experiences of discrimination and ior, once psychological distress and alcohol and/or drug use
sexual risk behavior, psychological distress and alcohol before/during sex were controlled. Results of recent anal-
and/or drug use before/during last sex, were associated yses of the role of internalized homophobia and sexual risk
with the outcome. Including these factors in the model with behavior have been mixed. One study reported that inter-
the full sample attenuated the relationship between sexual nalized homophobia mediated the relationship between
orientation-based discrimination and acquisition risk heterosexist discrimination and depression among a sample
behavior, as evidenced by the diminishment of the odds of MSM [66]. A recent meta-analysis, which assessed

123
AIDS Behav (2015) 19:257–269 265

Table 5 Sociodemographic, sexual behavior, and identity- and attachment-related characteristics and HIV acquisitiona risk behavior among
African American/Black and Latino MSM, M2MNYC, N = 498
Characteristics Total Acquisition risk (N = 498) p value
N (%) Yes No

Recruitment N (%) N (%) NS

Venue 273 (54.8) 27 (50.9) 246 (55.3)
Online 225 (45.2) 26 (49.1) 199 (44.7)
Sociodemographic
Age (Mean, SD) 29.5 (9.1) 28.7 (9.7) 29.6 (9.1) NS
Age, categorized NS
18–24 176 (35.3) 19 (35.9) 157 (35.3)
25–29 142 (28.5) 19 (35.9) 123 (27.6)
30–39 104 (20.9) 10 (18.9) 94 (21.1)
40? 76 (15.3) 5 (9.4) 71 (16.0)
Race/ethnicity NS
Black 208 (41.8) 22 (41.5) 186 (41.8)
Hispanic 290 (58.2) 31 (58.5) 259 (58.2)
Education NS
Less than high school graduate 31 (6.2) 5 (9.4) 26 (5.8)
High school graduate 72 (14.5) 12 (22.6) 60 (13.5)
Some college 208 (41.8) 17 (32.1) 191 (42.9)
College graduate or more 187 (37.6) 19 (35.9) 168 (37.8)
Employment NS
Working full-time 188 (37.8) 20 (37.7) 168 (37.8)
Working part-time 133 (26.7) 11 (20.8) 122 (27.4)
Not working, looking/not working, 158 (31.7) 22 (41.5) 136 (30.6)
not looking/temporarily laid off/retired
Working off the book/other 19 (3.8) 0 (0.0) 19 (4.3)
Personal income NS
\$10,000 154 (31.7) 24 (45.3) 130 (30.0)
$10,000–39,999 211 (43.4) 19 (35.9) 192 (44.3)
$40,000–59,999 72 (14.8) 4 (7.6) 68 (15.7)
$60,000? 49 (10.1) 6 (11.3) 43 (9.9)
Financial insecurity
Not enough $ for rent, food, or utilities 265 (53.5) 30 (56.6) 235 (53.2) NS
Not enough $ for social activity 372 (74.9) 42 (79.3) 330 (74.3) NS
Partnership Status
Married or registered domestic partner 16 (3.2) 3 (5.7) 13 (2.9) NS
Sexual identity NS
Gay, homosexual, queer, same gender loving, etc. 407 (81.7) 46 (86.8) 361 (81.1)
Bisexual 70 (14.1) 6 (11.3) 64 (14.4)
Straight, heterosexual 5 (1.0) 0 (0.0) 5 (1.1)
Unsure or questioning/other/missing 16 (3.2) 1 (1.9) 15 (3.4)
Mean (SD)
‘‘Outness’’ (Mean, SD) 7.8 (2.5) 7.9 (2.5) 7.8 (2.5) NS
Racial/ethnic identity (Mean, SD) 3.1 (0.5) 3.0 (0.5) 3.1 (0.5) NS
Gay community attachment (Mean, SD) 3.1 (0.5) 3.0 (0.6) 3.1 (0.5) NS
Condom use-related
Condom use self efficacy (Mean, SD) 4.3 (0.6) 3.9 (0.6) 4.3 (0.6) \.001

123
266 AIDS Behav (2015) 19:257–269

Table 5 continued
Characteristics Total Acquisition risk (N = 498) p value
N (%) Yes No

Perceived peer condom use norms (Mean, SD) 3.0 (0.5) 3.0 (0.5) 3.0 (0.6) NS
a
HIV acquisition risk behavior was coded as 1 among HIV-negative participants who had unprotected receptive anal sex with any type (e.g.,
primary, casual, etc.) of HIV-positive or unknown HIV status male sex partner; the remaining HIV-negative participants were coded as 0

Table 6 Unadjusted associations among sexual orientation- (SO) and bias, would not have provided positive responses to these
race-based discrimination and HIV acquisitiona risk behavior among questions. In addition, the measure of race/ethnicity-based
African American/Black and Latino MSM, M2MNYC, N = 498
discrimination did not assess individually the various
Characteristics Total Acquisition p value domains where discrimination occurs (e.g., home, school,
Risk (N = 498) work, etc.) or varying forms or levels of discrimination
Either home or social N (%) (e.g., micro-aggressions), which resulted in a less sensitive
neighborhood measure [69]. Third, because of low prevalence of sexual
None 412 (82.7) ref 0.297 transmission risk behavior among the HIV-positive par-
Sexual orientation-based 29 (5.8) 2.24 0.86, 5.80 ticipants in our sample, we were unable to model relations
only between discrimination and this outcome. Finally, we have
Race-based only 29 (5.8) 0.64 0.15, 2.77 used the phrase ‘‘unprotected’’ sex to denote sex without a
Both 28 (5.6) 0.66 0.15, 2.88 condom; we recognize that some HIV-negative MSM may
a
HIV acquisition risk behavior was coded as 1 among HIV-negative have been having sex with HIV-positive partners whose
participants who had unprotected receptive anal sex with any type viral loads were undetectable, significantly reducing
(e.g., primary, casual, etc.) of HIV-positive or unknown HIV status acquisition risk. Alternatively, the participant may have
male sex partner; the remaining HIV-negative participants were been using pre-exposure prophylaxis (PrEP) ; however, at
coded as 0
the time of data collection, PrEP was not approved for use
and was not commonly available.
relations among internalized homophobia and internalizing
mental health problems (e.g., anxiety and depression),
concluded that the association may be decreasing over Conclusions
time, as the authors found a moderating effect of the year
of publication of the study [67]. We also assessed the Our analysis builds on the work of several recent studies
impact of attachment to the gay community and racial/ that have examined the role of social discrimination in
ethnic identity on the relationship between social dis- sexual risk behavior. We have extended this work by
crimination and risk behavior, but did not find evidence of examining different forms of social discrimination in
any protective effect of either factor. This finding contrasts combination and separately. We have also explored whe-
with the work of O’Donnell and colleagues (2002) and ther place of experienced discrimination, either the home
Chng and Geliga-Vargas (2000) among Latino men [42, or social neighborhood, related to sexual risk behavior,
43]. However other recent studies have also not found which it did not. We stratified our sample by the risk
ethnic or racial identity or community affiliation to be behavior outcome, understanding that correlates of acqui-
associated with lower sexual risk behavior among MSM sition risk behavior among HIV-negative men are different
[35, 68]. from transmission risk behaviors among HIV-positive and
This analysis has several limitations that must be taken unknown status men. In addition, we examined the roles of
into account when considering the results. First, as a cross- internalized homophobia, psychological distress and alco-
sectional study, it is impossible to infer causality from the hol and/or drug use before/during sex, factors that may link
correlations reported here; a prospective cohort study discrimination and sexual HIV acquisition risk behavior.
would be required to properly identify mediation and/or Finally, we included in our models individual-level
causal relations among the factors studied. Second, the potential predictors of condom use, in order to properly
assessment of discrimination relied on participants evalu- specify the models.
ating the discriminatory treatment they received as being The finding that sexual orientation-based discrimination
due to their sexual orientation and/or race/ethnicity; par- in home or social neighborhoods is associated with sexual
ticipants who were discriminated against, but who did not HIV acquisition risk behavior is consistent with previous
attribute the behavior to racial or sexual orientation-based research and suggests that further effort should be

123
AIDS Behav (2015) 19:257–269 267

expended to address the root causes of sexual orientation- 9. Egan JE, Frye V, Kurtz SP, Latkin C, Chen M, Tobin K, et al.
based discrimination, such as homophobia and HIV/AIDS Migration, neighborhoods, and networks: approaches to under-
standing how urban environmental conditions affect syndemic
stigma. Existing interventions to decrease homophobia and adverse health outcomes among gay, bisexual and other men who
HIV/AIDS stigma have provided information and educa- have sex with men. AIDS Behav. 2011;15(Suppl 1):S35–50.
tion, worked to change attitudes and values, and increased 10. Frye V, Koblin B, Chin J, Beard J, Blaney S, Halkitis P, et al.
contact with people living with HIV/AIDS, activities that Neighbourhood-level correlates of consistent condom use among
men who have sex with men: a multi-level analysis. AIDS Behav.
reduce stereotyping, prejudice, and acts of sexual orienta- 2010;14(4):974–85.
tion-based discrimination [70, 71]. However, most of these 11. Kelly BC, Carpiano RM, Easterbrook A, Parsons JT. Sex and the
interventions have been conducted at the individual level, community: the implications of neighbourhoods and social net-
with a meta-analysis of HIV prevention interventions for works for sexual risk behaviours among urban gay men. Sociol
Health Illn. 2012;34(7):1085–102.
MSM reporting no community-level interventions focused 12. Buttram ME, Kurtz SP. Risk and protective factors associated with
on social discrimination, such as sexual orientation-based gay neighborhood residence. Am J Mens Health. 2013;7(2):110–8.
discrimination and/or homophobia [72]. Further longitu- 13. Millett GA, Peterson JL, Flores SA, Hart TA, Jeffries WL, Wilson
dinal research is needed to determine whether reductions in PA, et al. Comparisons of disparities and risks of HIV infection in
black and other men who have sex with men in Canada, UK, and
sexual HIV risk behavior will occur as a result of USA: a meta-analysis. Lancet. 2012;380(9839):341–8.
decreasing exposure to sexual orientation-based discrimi- 14. Smedley BD. The lived experience of race and its health conse-
nation among MSM, a critical social goal independent of quences. Am J Public Health. 2012;102(5):933–5.
its role in the HIV epidemic. 15. Crenshaw K. Mapping the margins: intersectionality, identity
politics, and violence against women of color. Stanford Law Rev.
1991;43(6):1241–99.
Acknowledgments The authors acknowledge the assistance of 16. Stall R, Mills TC, Williamson J, Hart T, Greenwood G, Paul J,
Krista Goodman, DaShawn Usher, Travis Cronin, Jermaine et al. Association of co-occurring psychosocial health problems
McCrossin, John Bonelli, Geneva Ortiz, Kiwan Stewart, Kimberly and increased vulnerability to HIV/AIDS among urban men who
Morales, Evelyn Rivera, Blaz Bush, Heriberto Oquendo, Jr., Michael have sex with men. Am J Public Health. 2003;93(6):939–42.
Hernandez, and Damon Jacobs. We also thank the individuals who 17. Singer M. Introduction to syndemics: a critial systems approach to
gave their time and efforts to participate in this study. This study was public and community health. San Francisco: Jossey-Bass; 2009.
supported by National Institute of Child Health and Human Devel- 18. Westbrook L, Schilt K. Doing gender, determining gender: trans-
opment R01 HD059729 to Dr. Beryl Koblin of the New York Blood gender people, gender panics, and the maintenance of the sex/
Center. Dr. Victoria Frye’s work was also supported through a gender/sexuality system. Gender and Society. 2014;28(1):32–57.
National Institute on Drug Abuse mentored career development award 19. Namaste VK. Genderbashing: sexuality gender, and the regula-
(K01 DA-020774). tion of public space. In: Stryker S, Whittle S, editors. The
transgender studies reader. London: Routledge; 2006.
20. Wilchins RA. Queer theory, gender theory: an instant primer. Los
Angeles: Alyson Publications; 2004.
References 21. Massey D, Denton NA. The dimensions of residential segrega-
tion. Soc Forces. 1998;67:281–315.
1. Centers for Disease Control and Prevention. HIV Surveillance 22. Williams DR, Mohammed SA. Discrimination and racial dis-
Report, 2009; vol. 21. 2011 parities in health: evidence and needed research. J Behav Med.
2. Prejean J, Song R, Hernandez A, Ziebell R, Green T, Walker F, 2009;32(1):20–47.
et al. Estimated HIV incidence in the United States, 2006–2009. 23. Kessler RC, Mickelson KD, Williams DR. The prevalence, distri-
PLoS One. 2011;6(8):e17502. bution, and mental health correlates of perceived discrimination in
3. New York City Deparment of Health. HIV Epidemiology and the United States. J Health Soc Behav. 1999;40(3):208–30.
Field Services Semiannual Report. 2012; 7(1). 24. Meyer IH. Minority stress and mental health in gay men. J Health
4. Millett GA, Flores SA, Peterson JL, Bakeman R. Explaining Soc Behav. 1995;36:38–56.
disparities in HIV infection among black and white men who 25. DeLongis A, Lazarus RS, Folkman S. The impact of daily stress
have sex with men: a meta-analysis of HIV risk behaviors. AIDS. on health and mood: psychological and social resources as
2007;21(15):2083–91. mediators. J Pers Soc Psychol. 1988;54(3):486–95.
5. Harawa NT, Greenland S, Bingham TA, Johnson DF, Cochran 26. Lazarus RS, Folkman S. Stress, appraisal. New York: Coping;
SD, Cunningham WE, et al. Associations of race/ethnicity with 1984.
HIV prevalence and HIV-related behaviors among young men 27. Meyer I. Prejudice, social stress, and mental health in lesbian, gay
who have sex with men in 7 urban centers in the United States. and bisexual populations: conceptual issues and research evidence.
J Acquir Immune Defic Syndr. 2004;35(5):526–36. Psychol Bull. 2003;129(5):674–97.
6. Koblin BA, Husnik MJ, Colfax G, Huang Y, Madison M, Mayer 28. Gottheil E. Stress and addiction. New York: Brunner Mazel;
K, et al. Risk factors for HIV infection among men who have sex 1987.
with men. AIDS. 2006;20(5):731–9. 29. Chae DH, Krieger N, Bennett GG, Lindsey JC, Stoddard AM,
7. Koblin BA, Mayer KH, Eshleman SH, Wang L, Mannheimer S, Barbeau EM. Implications of discrimination based on sexuality,
del Rio C, et al. Correlates of HIV acquisition in a cohort of black gender, and race/ethnicity for psychological distress among
men who have sex with men in the United States: HIV prevention working-class sexual minorities: the United for Health Study,
trials network (HPTN) 061. PLoS One. 2013;8(7):e70413. 2003–2004. Int J Health Serv. 2010;40(4):589–608.
8. Latkin CA, German D, Vlahov D, Galea S. Neighborhoods and 30. Wong CF, Weiss G, Ayala G, Kipke MD. Harassment, discrim-
HIV: a social ecological approach to prevention and care. Am ination, violence, and illicit drug use among young men who have
Psychol. 2013;68(4):210–24. sex with men. AIDS Educ Prev. 2010;22(4):286–98.

123
268 AIDS Behav (2015) 19:257–269

31. Mizuno Y, Borkowf C, Millett GA, Bingham T, Ayala G, Stueve A. 49. Koblin B, Egan JE, Rundle A, Quinn T, Tieu HV, et al. Methods
Homophobia and racism experienced by Latino men who have sex to measure the impact of home, social, and sexual neighborhoods
with men in the United States: correlates of exposure and associ- of urban gay, bisexual, and other men who have sex with men.
ations with HIV risk behaviors. AIDS Behav. 2012;16(3):724–35. PLoS Med. 2013;8(10):e75878.
32. Fields EL, Bogart LM, Galvan FH, Wagner GJ, Klein DJ, 50. Mackellar DA, Gallagher KM, Finlayson T, Sanchez T, Lansky A,
Schuster MA. Association of discrimination-related trauma with Sullivan PS. Surveillance of HIV risk and prevention behaviors of
sexual risk among HIV-positive African American men who have men who have sex with men–a national application of venue-based,
sex with men. Am J Public Health. 2013;103(5):875–80. time-space sampling. Public Health Rep. 2007;122(Suppl 1):39–47.
33. Ayala G, Bingham T, Kim J, Wheeler DP, Millett GA. Modeling 51. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief
the impact of social discrimination and financial hardship on the depression severity measure. J Gen Intern Med. 2001;16(9):606–13.
sexual risk of HIV among Latino and Black men who have sex 52. Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure
with men. Am J Public Health. 2012;102(Suppl 2):S242–9. for assessing generalized anxiety disorder: the GAD-7. Arch
doi:10.2105/AJPH.2011.300641. Intern Med. 2006;166(10):1092–7.
34. Malebranche DJ, Gvetadze R, Millett GA, Sutton MY. The 53. Ross MW, Rosser BRS, Bauer GR, Bockting WO, Robinson
relationship between gender role conflict and condom use among BBE, Rugg DL, et al. Drug use, unsafe sexual behavior, and
black MSM. AIDS Behav. 2012;16(7):2051–61. internalized homonegativity in men who have sex with men.
35. Jeffries WL, Marks G, Lauby J, Murrill CS, Millett GA. AIDS Behav. 2001;5(1):97–103.
Homophobia is associated with sexual behavior that increases 54. Rowen CJ, Malcolm JP. Correlates of internalized homophobia
risk of acquiring and transmitting HIV Infection among black and homosexual identity formation in a sample of gay men.
men who have sex with men. AIDS Behav. 2012;17(4):1442–53. J Homosex. 2002;43(2):77–92.
36. Karlsen S, Nazroo JY. Relation between racial discrimination, 55. Phinney J. The multigroup ethnic identity measure: a new scale
social class, and health among ethnic minority groups. Am J for use with adolescents and young adults from diverse groups.
Public Health. 2002;92(4):624–31. J Adolesc Res. 1992;7:156–76.
37. Nazroo JY. The structuring of ethnic inequalities in health: eco- 56. Bradford LJ, Beck KH. Development and validation of the con-
nomic position, racial discrimination, and racism. Am J Public dom self-efficacy scale for college students. J Am Coll Health.
Health. 2003;93(2):277–84. 1991;39(2):219–25.
38. Egan JE, Frye V, Kurtz SP, Latkin C, Chen M, Tobin K, et al. 57. Choi KH, Han CS, Hudes ES, Kegeles S. Unprotected sex and
Migration, neighborhoods, and networks: approaches to under- associated risk factors among young Asian and Pacific Islander
standing how urban environmental conditions affect syndemic men who have sex with men. AIDS Educ Prev. 2002;14(6):472–81.
adverse health outcomes among gay, bisexual and other men who 58. DiIorio C, Maibach E, O’Leary A, Sanderson CA, Celentano D.
have sex with men. AIDS Behav. 2011;15(Suppl 1):S35–50. Measurement of condom use self-efficacy and outcome expec-
39. Diggs M. Surveying the intersection: pathology, secrecy, and the tancies in a geographically diverse group of STD patients. AIDS
discourses of racial and sexual identity. J Homosex. 1993;26(2–3): Educ Prev. 1997;9(1):1–13.
1–19. 59. Jones KT, Johnson WD, Wheeler DP, Gray P, Foust E, Gaiter J.
40. Diaz RM, Ayala G, Bein E. Sexual risk as an outcome of social Nonsupportive peer norms and incarceration as HIV risk corre-
oppression: data from a probability sample of Latino gay men in lates for young black men who have sex with men. AIDS Behav.
three US cities. Cultur Divers Ethnic Minor Psychol. 2004;10(3): 2008;12(1):41–50.
255–67. 60. Koblin BA, Bonner S, Powell B, Metralexis P, Egan JE, Patterson
41. Marks G, Crepaz N, Senterfitt JW, Janssen RS. Meta-analysis of high- J, et al. A randomized trial of a behavioral intervention for black
risk sexual behavior in persons aware and unaware they are infected MSM: the DiSH study. AIDS. 2012;26(4):483–8.
with HIV in the United States: implications for HIV prevention 61. Jeffries WL, Marks G, Lauby J, Murrill CS, Millett GA.
programs. J Acquir Immune Defic Syndr. 2005;39(4):446–53. Homophobia is associated with sexual behavior that increases
42. O’Donnell L, Agronick G, San DA, Duran R, Myint U, Stueve A. risk of acquiring and transmitting HIV infection among black
Ethnic and gay community attachments and sexual risk behaviors men who have sex with men. AIDS Behav. 2013;17(4):1442–53.
among urban Latino young men who have sex with men. AIDS 62. Fingerhut AW, Peplau LA, Gable SL. Identity, minority stress
Educ Prev. 2002;14(6):457–71. and psychological well-being among gay men and lesbians.
43. Chng CL, Geliga-Vargas J. Ethnic identity, gay identity, sexual Psychol Sex. 2010;1(2):101–14.
sensation seeking and HIV risk taking among multiethnic men 63. McCabe SE, Bostwick WB, Hughes TL, West BT, Boyd CJ. The
who have sex with men. AIDS Educ Prev. 2000;12(4):326–39. relationship between discrimination and substance use disorders
44. Flores SA, Mansergh G, Marks G, Guzman R, Colfax G. Gay among lesbian, gay, and bisexual adults in the United States. Am
identity-related factors and sexual risk among men who have sex J Public Health. 2010;100(10):1946–52.
with men in San Francisco. AIDS Educ Prev. 2009;21(2):91–103. 64. Morrison MA. Psychological health correlates of perceived dis-
45. Rosario M, Schrimshaw EW, Hunter J. A model of sexual risk crimination among canadian Gay men and Lesbian women. Can J
behaviors among young gay and bisexual men: longitudinal Commun Ment Health. 2011;30(2):81–98.
associations of mental health, substance abuse, sexual abuse, and 65. Feinstein BA, Goldfried MR, Davila J. The relationship between
the coming-out process. AIDS Educ Prev. 2006;18(5):444–60. experiences of discrimination and mental health among lesbians
46. White D, Stephenson R. Identity formation, outness, and sexual and gay men: an examination of internalized homonegativity and
risk among gay and bisexual men. Am J Mens Health. rejection sensitivity as potential mechanisms. J Consult Clin
2014;8(2):98–109. Psychol. 2012;80(5):917–27.
47. Klein H. Condom use self-efficacy and HIV risk practices among 66. Szymanski DM, Ikizler AS. Internalized heterosexism as a
men who use the internet to find male partners for unprotected mediator in the relationship between gender role conflict, het-
sex. Am J Mens Health. 2014;8(3):190–204. erosexist discrimination, and depression among sexual minority
48. Bauermeister JA, Hickok AM, Meadowbrooke C, Veinot T, men. Psychol Men Masc. 2013;14(2):211.
Loveluck J. Self-efficacy among young men who have sex with 67. Newcomb ME, Mustanski B. Internalized homophobia and
men: an exploratory analysis of HIV/AIDS risk behaviors across internalizing mental health problems: a meta-analytic review.
partner types. AIDS Behav. 2014;18(1):69–77. Clin Psychol Rev. 2010;30(8):1019–29.

123
AIDS Behav (2015) 19:257–269 269

68. Warren JC, Fernandez MI, Harper GW, Hidalgo MA, Jamil OB, 71. Mahajan AP, Sayles JN, Patel VA, Remien RH, Sawires SR,
Torres RS. Predictors of unprotected sex among young sexually Ortiz DJ, et al. Stigma in the HIV/AIDS epidemic: a review of the
active African American, Hispanic, and White MSM: the impor- literature and recommendations for the way forward. AIDS.
tance of ethnicity and culture. AIDS Behav. 2008;12(3):459–68. 2008;22(Suppl 2):S67–79.
69. Krieger N, Smith K, Naishadham D, Hartman C, Barbeau EM. 72. Herbst JH, Beeker C, Mathew A, McNally T, Passin WF, Kay
Experiences of discrimination: validity and reliability of a self- LS, et al. The effectiveness of individual-, group-, and commu-
report measure for population health research on racism and nity-level HIV behavioral risk-reduction interventions for adult
health. Soc Sci Med. 2005;61(7):1576–96. men who have sex with men: a systematic review. Am J Prev
70. Klein SJ, Karchner WD, O’Connell DA. Interventions to prevent Med. 2007;32(4 Suppl):S38–67.
HIV-related stigma and discrimination: findings and recommenda-
tions for public health practice. J Public Health Manage Pract.
2002;8(6):44–53.

123
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.