Paper 11 Ok

C 2005) AIDS and Behavior, Vol. 9, No. 2, June 2005 (
DOI: 10.1007/s10461-005-3897-8

Partner Awareness of the Serostatus of HIV-Seropositive Men Who Have Sex With Men: Impact on Unprotected Sexual Behavior Trevor A. Hart,1,5 Richard J. Wolitski,2 David W. Purcell,2 Jeffrey T. Parsons,3 4 ´ Cynthia A. Gomez, and the Seropositive Urban Men’s Study Team

Prior research has provided conflicting evidence about the association between partner awareness of an HIV-seropositive person’s serostatus and HIV transmission behavior via unprotected intercourse. The current study examined partner awareness of participant HIVseropositive status and sexual behavior in a multiethnic sample of HIV-seropositive men who have sex with men. Most HIV-seropositive men reported that their primary partners are aware, and most reported that at least some non-primary partners are aware the participant was HIV-seropositive before first having sex. Partner awareness of participant HIVserostatus was related to unprotected sexual behavior during the past 3 months in a non-linear fashion, as men with partners who were inconsistently aware had higher rates of unprotected receptive anal intercourse than men with partners who were consistently aware or consistently unaware. Men with partners who were inconsistently aware also had higher rates of insertive oral intercourse than men with partners who were consistently aware. However, there were no differences in HIV transmission risk behavior between men with partners who were consistently aware and men with partners who were consistently unaware a participant was HIV-seropositive. KEY WORDS: HIV; awareness; seropositive; men who have sex with men.

INTRODUCTION

Stein and Samet, 1999), it is also possible that if safer sex is practiced then partner awareness would not further reduce the risk of HIV transmission. From a public health perspective, the literature has provided conflicting evidence about the association between partner awareness that an individual is seropositive and practicing unprotected intercourse. Some studies have found that partner awareness was associated with increased unprotected intercourse (Kalichman and Nachimson, 1999; Niccolai et al., 1999), while others have not (Marks and Crepaz, 2001; Ostrow et al., 1989; Wolitski et al., 1998). The vast majority of the literature on partner awareness has examined the role of overt disclosure of serostatus of the HIVseropositive partner. The relation of overt disclosure and unprotected intercourse has been examined during a specific sexual encounter (e.g., Marks and Crepaz, 2001) or with one partner (e.g., De Rosa and

A continuing debate exists as to the importance of awareness of serostatus among sex partners of HIV-seropositive persons and its relationship to sexual risk behaviors. Although some have argued that HIV-seropositive individuals always need to ensure that potential partners are aware that they are having sex with an HIV-seropositive individual so that noninfected partners can make informed decisions about their risk of HIV infection (Marks and Crepaz, 2001; 1 York

University. for Disease Control and Prevention. 3 City University of New York. 4 University of California at San Francisco. 5 Correspondence should be directed to Trevor A. Hart, Department of Psychology, York University, 4700 Keele St., Toronto, ON M3J 1P3: Canada; e-mail: [email protected]. 2 Centers

155 C 2005 Springer Science+Business Media, Inc. 1090-7165/05/0600-0155/0


156

´ Hart, Wolitski, Purcell, Parsons, Gomez, and the Seropositive Urban Men’s Study Team

Marks, 1998). Other studies have employed a retrospective self-report across a specified timeframe, and assessed differences between HIV-seropositive individuals who disclosed to all partners or not (e.g., Stein et al., 1998) or those who disclosed to at least one partner or none of their partners (e.g., Kalichman and Nachimson, 1999; Wolitski et al., 1998). Although each of these methods have added significantly to understanding of disclosure, none of them allows comparison of respondents who have partners who are inconsistently aware with respondents who report that all of their partners or none of their partners were aware that the respondent was HIV-seropositive prior to risky sexual behavior. Several behavioral, psychological, and medical variables have been examined as predictors of HIV disclosure. Contextual behavioral variables previously associated with non-disclosure among men who have sex with men (MSM) include greater number of partners (Niccolai et al., 1999; Wolitski et al., 1998), and having anonymous or casual sex partners (Semple et al., 2000). Qualitative data on the context in which disclosure occurs also suggest a possible association between non-disclosure and meeting partners at sex venues such as bathhouses or public sex venues (Stirratt, 2005). Having partners in the context of a less emotionally intimate context may therefore be associated with less disclosure and/or partner awareness of serostatus. Psychological variables have included discomfort with a gay sexual identity (Simoni et al., 1997; Perry et al., 1994), low disclosure self-efficacy (Kalichman and Nachimson, 1999), and low disclosure intention (Stirratt, 2005). Medical variables associated with non-disclosure have included having a non-symptomatic HIV infection and lower number of years since being diagnosed (Crepaz and Marks, 2003; Mansergh et al., 1995; Marks et al., 1992; Simoni et al., 1997). Most studies (e.g., Zea et al., 2003) have examined partner awareness by examining verbal disclosure. However, qualitative data suggests that many HIV-seropositive MSM report their sexual partners become aware by other means, such as attending social events or other programs for HIV-seropositive individuals also attended by the participant (Stirratt, 2005). Verbal disclosure may therefore fail to capture all instances in which an individual has been made aware that a partner is HIV-seropositive. The current study attempted to answer four primary research questions. First, the study extended the findings of previous disclosure research by examining awareness of HIV serostatus among sexual

partners and sexual behavior in a multiethnic sample of HIV-seropositive MSM. Second, the current study sought to examine partner awareness and HIV transmission risk with non-primary partners. HIV transmission risk behaviors were defined as unprotected insertive and receptive anal intercourse, and unprotected insertive oral intercourse with HIV-negative or unknown serostatus partners. Men were categorized by overall pattern of partner awareness into three groups: men with all HIV-negative or unknown serostatus non-primary partners who were aware (consistently aware), men with no HIV-negative or unknown serostatus non-primary partners who were aware (consistently unaware), and men with some but not all partners who were aware (inconsistently aware) that the participant was HIV-seropositive before sex. Third, given the lack of literature on sexual partners’ inconsistent versus consistent awareness or non-awareness of serostatus of HIV-seropositive MSM, it was of interest to explore differences among groups regarding variables previously found to be associated with greater awareness: a) psychological variables, such as disclosure self-efficacy, disclosure intention, comfort discussing HIV-serostatus, comfort with a gay sexual identity, b) contextual behavioral variables including lower number of male partners, lower number of HIV-negative or unknown serostatus partners, having less than all non-primary partners being one-night stands, and not attending relatively anonymous sex venues (e.g., bathhouses, public sex venues) to search for sex partners in the past 3 months, and c) medical variables including having an AIDS diagnosis and greater number of years since HIV diagnosis. Fourth, in order to examine if awareness was an independent predictor of HIV transmission risk, it was of interest to explore if patterns of awareness would predict HIV transmission risk even when controlling for psychological and behavioral correlates of awareness and sexual behavior. METHODS Participants The sample consisted of 456 participants recruited from June 1997 to September 1998 in New York City and San Francisco for the Seropositive Urban Men’s Study (Purcell et al., 2001; Wolitski et al., 2001). This study was designed to examine factors associated with safer sex behaviors among HIVseropositive men who have sex with men. Eligibility

Partner Awareness of HIV Serostatus requirements consisted of being at least 18 years of age, male (or having male genitalia if self-identified as transgender), self-identifying as HIV-positive, reporting having sex with another man in the past year, and living in the study catchment area. Recruitment quotas resulted in approximately equal numbers of African-Americans (29%), Latinos (24%), and Caucasians (30%), with the remainder of the sample consisting of Asian or Pacific Islanders (6%), men of mixed race (8%), or others (3%). Age ranged from 20 to 66 years with a mean of 37.4 (SD = 8.0). Approximately 61% were from New York and 39% from San Francisco. Regarding sexual orientation, 83% identified as gay, homosexual, or queer, 10% identified as bisexual, 1% identified as heterosexual, and 4% were unsure or reported none of the above. The vast majority (98%) identified as male, with the remainder identifying as transgender. The sample was primarily low-income, with 44% making less than $10,000 a year, 25% making $10,000–19,999 a year, and the remainder (31%) making $20,000 or more. Only 20% were employed full-time, and 13% were employed part-time. Approximately 35% were receiving unemployment due to disability, 4% were unemployed students, and 28% were unemployed for other reasons. Approximately 7% of the men did not complete high school, 19% completed high school, 38% attended some college, 24% completed college, and 11% received a graduate degree. The mean time that participants knew of their HIV-positive status was 6.8 years (SD = 4.0), and 44% of the men had been diagnosed with AIDS. The vast majority (97%) of men reported knowing their CD4 count (mean CD4 count = 396, SD = 274). Most (92%) had their viral load tested, with 42% reporting an “undetectable” viral load. Most (79%) were taking HIV medications, with 81% of these taking a regimen including a protease inhibitor.

157 icant other, or life partner.” Participants were then asked about the serostatus of the primary partner, and if the primary partner “knew your HIV status before you had sex together for the first time.” Regarding non-primary partners, participants were asked how many HIV-seropositive, HIV-seronegative, and unknown serostatus partners they had in the past 3 months. A partner’s serostatus could be unknown to the respondent because a) it was unknown to the partner or b) the partner did not report his serostatus to the respondent. Participants were then asked separately by partner serostatus how many partners of that serostatus were aware of the participants’ serostatus before having sex for the first time. For participants reporting non-primary partners, participant responses were categorized into three groups: those for whom all non-primary HIV-negative or unknown serostatus partners were aware (consistently aware), those for whom some but not all non-primary HIV-negative or unknown serostatus partners were aware (inconsistently aware), and those for whom none of their non-primary HIV-negative or unknown serostatus partners were aware (unaware) the participant was HIV-seropositive. Disclosure Self-Efficacy Participants were asked how much they agreed with statements regarding their perceived ability to disclose their HIV-seropositive status to sex partners across 5 situations (e.g., “I can disclose my HIV status before having sex even to a really hot new sex partner.”). Participants indicated agreement using a 5-point Likert-type scale (1 = absolutely sure I cannot, 5 = absolutely sure I can). The measure was internally consistent, α = .90. Disclosure Intention

Measures Partner Awareness of Respondents’ HIV-Positive Serostatus Participants were asked about partner awareness of participant HIV-positive serostatus among primary sexual partners and non-primary partners with whom the participant had sex in the last 3 months. Regarding primary partners, participants were asked if they had a primary partner “that is, a partner you would call your boyfriend, spouse, signif-

Participants were asked how much they agreed with the following statement: “I plan on telling my HIV status to all of my new partners before we have sex.” Participants answered this question using a 5point Likert-type scale (1 = strongly disagree, 5 = strongly agree). Preferences for Nonverbal Versus Verbal Disclosure Participants were asked if they “prefer to tell new sex partners about your HIV status using verbal

158

´ Hart, Wolitski, Purcell, Parsons, Gomez, and the Seropositive Urban Men’s Study Team

(i.e., tell them directly) or non-verbal communication (i.e., tell them without words).” Comfort Talking About HIV Serostatus Participants were asked how comfortable they were talking to close friends, casual sex partners, or health care providers about their HIV serostatus. Items were treated separately to examine differences in disclosure behavior for different types of relationships. Participants indicated their comfort using a 5point Likert-type scale (1 = very uncomfortable, 5 = very comfortable). Comfort Regarding Gay Identity Comfort regarding gay sexuality among this MSM sample was operationalized via gay selfidentification and internalized homophobia. Regarding gay self-identification, participants were asked if they self-identified as 1) gay, homosexual, or queer, 2) bisexual, 3) heterosexual, or 4) unsure of their sexual orientation. The internalized homophobia measure contains 5-items assessing degree of discomfort with one’s own homosexuality. Most items are reverse coded (e.g., “I feel extremely comfortable about being sexually attracted to men”). The scale was adapted from Ross and Rosser’s (1996) internalized homophobia scale and demonstrates acceptable internal consistency (α = .77). Attended Sex Venues to Meet Sex Partners Participants were asked if they went to sex clubs, bathhouses or public cruising areas where men look for sex (e.g., parks) to meet new sex partners in the past 3 months. One Night Stands Participants were asked what proportion of their non-primary partners were one-time partners (i.e., one night stands) using a 5-point Likert-type scale (1 = none, 5 = all or almost all). Because the vast majority (89%) of men who had non-primary partners reported having had one-night stands and 36% reported all or almost all non-primary partners were one-night stands, responses were dichotomized to indicate whether all/almost all partners were one night stands or not.

Social Desirability Response Bias Participants completed the Marlowe-Crowne Social Desirability Scale-Short Form (Crowne and Marlowe, 1964; Reynolds, 1992) to assess for social desirability reporting bias. This instrument consists of 13 items such as “I’m always willing to admit when I make a mistake.” The MarloweCrowne Social Desirability Scale has demonstrated high test-retest reliability and internal consistency in the general population (Crowne and Marlowe, 1964; Reynolds, 1992) and high internal consistency in a study of unprotected sexual behavior among MSM (DiFranceisco et al., 1998). It demonstrated similar internal consistency in the present sample, α = .71.

Sexual Behavior and HIV Transmission Risk Behavior Participants were asked about sexual activities in the past 3 months with their primary sex partner and non-primary partners who were HIVseropositive, HIV-seronegative, or of unknown serostatus. Questions included the frequency of insertive anal intercourse and receptive anal intercourse. HIV transmission risk behavior was assessed by asking participants about the number of times they had unprotected insertive and receptive anal intercourse with partners of HIV-negative or unknown serostatus in the last 3 months. Participants were also asked about the number of times they had unprotected insertive oral intercourse in the last 3 months. To be consistent with the previous literature on disclosure and unprotected sex, and because any act of unprotected intercourse between serodiscordant partners is a potential risk for HIV transmission and is therefore of public health concern, number of occasions was then dichotomized to indicate presence or absence of an unprotected behavior (e.g., unprotected receptive anal intercourse). This dichotomization also served to avoid violating assumptions regarding normal distribution for linear regressions. The proportion of each group engaging in an unprotected sexual behavior was calculated by dividing the number of participants engaging in an unprotected behavior by the total number of participants engaging in any instance of that behavior (e.g., number engaging in unprotected receptive anal intercourse divided by number engaging in any receptive anal intercourse).

Partner Awareness of HIV Serostatus Setting and Procedure

159

Sampling quotas were established to ensure the representation of key ethnic/racial groups recruited from three types of community settings: AIDS service organizations, mainstream gay venues such as gay bookstores, bars, and gay pride events, and public sex environments (e.g., bathhouses, outdoor cruising areas, parks). Participants were actively recruited by field staff who approached potential participants in community venues and passively recruited through flyers placed in the venues and gay-oriented publications. Recruitment materials described the study and asked interested individuals to call a toll-free number to obtain additional information about the study. Interested participants were informed that the study examined a variety of issues relevant to HIVseropositive men, including feelings and experiences regarding sex, medical care, mental health and substance use. To eliminate the need for men to report their HIV status in the venue, potential participants were told, “If this does not apply to you, please give it to someone you know.” As a result, some (20%) men were referred to the study by friends and recruited into the study through snowball sampling. Men who met the eligibility criteria completed a paper-andpencil survey, which required approximately 1 hr on average to complete. An incentive of $30 was given to those completing the survey.

categorical variables, with follow-up chi-square analyses conducted when significant findings were found in the omnibus test. The continuity correction statistic was employed to determine significance in all 2 × 2 tables. Analyses of variance (ANOVAs) were conducted to investigate whether partner awareness group was associated with differences in continuous variables, with significant differences followed by Scheffe post hoc tests. Data regarding number of partners were highly positively skewed, and therefore were log-transformed (Kirk, 1968; Winer, 1971) using the formula log10 (x + 1) for all parametric analyses using number of partners as the dependent variable. Variables significantly associated with partner awareness group were then entered together as predictors in a multinomial logistic regression model with awareness group as the dependent variable. Hierarchical binary logistic regressions were computed with all significant correlates of partner awareness group entered on the first block and partner awareness group on the second block and an HIV transmission behavior as the dependent variable. All analyses were repeated selecting the 253 men who preferred to disclose verbally to examine whether differences among partner awareness groups differed as a function of preferences for verbal versus nonverbal disclosure. All differences found among partner awareness groups were significant at p < .05 unless stated otherwise.

Analyses

RESULTS

Chi-square analyses were conducted to investigate differences in proportion of seropositive, seronegative, and unknown serostatus primary partners who were aware the participant was HIVseropositive before first sexual contact. Proportion of seropositive, seronegative, and unknown serostatus non-primary partners aware was also calculated, with a generalized estimating equation (GEE) used to examine differences between rates of partner awareness 1) among primary partners versus non-primary partners and 2) among primary partners and nonprimary partners by partner serostatus (seropositive, seronegative, and unknown serostatus). Analyses then examined psychological, behavioral, and medical differences between participants with nonprimary HIV-negative or unknown serostatus partners who were consistently aware, inconsistently aware, and consistently unaware. Chi-square analyses were conducted to investigate whether partner awareness group was associated with differences in

Disclosure to Primary Versus Non-Primary Partners Thirty-nine percent of participants (n = 178) reported having sex with a primary partner in the last 3 months, compared to 80% (n = 363) who reported having sex with a non-primary partner. A quarter (n = 109) of the sample reported both primary and non-primary partners. Of the 178 participants who had sex with their primary partner in the last 3 months, more than three quarters (79%) reported their primary partner was aware they were HIV-seropositive before having sex with him for the first time. Primary partner awareness was not associated with the length of the relationship. The association between partner awareness and sexual practices with primary partners in the last 3 months could not be assessed because of the small number of men who reported their partner was not aware of the participant’s serostatus. Of the 363 participants who had sex with a non-primary partner in the last

160

´ Hart, Wolitski, Purcell, Parsons, Gomez, and the Seropositive Urban Men’s Study Team Table I. Mean Number of Partners and Proportion of Partners Aware By Serostatus of Partner (n = 456) Serostatus of partner Positive

Men with a primary partner (n) Proportion primary partners aware prior to first sexual contact Men with at least one non-primary partner (n) Number of non-primary partners (SD) Number of non-primary partners aware prior to first sexual contact (SD) Proportion non-partners aware prior to first sexual contact

Negative

Unknown

Total

91 72 15 178 87% 71% 50% 79% 198 146 285 363 2.46 (6.38) 1.48 (4.30) 8.82 (17.09) 12.65 (19.75) 2.02 (5.71) 1.07 (3.76) 2.30 (8.61) 5.19 (12.48) 84% 71% 26% 41%

Note. SD = standard deviation.

3 months, study participants reported that 41% of their non-primary partners were aware the participant was HIV-seropositive before having sex with the participant for the first time (see Table I). Analyses indicated that primary partners were more likely to be aware than non-primary partners, OR = 1.70, 95% CI = 1.49–2.53, and seropositive partners, OR = 2.07, 95% CI = 1.38–3.11, and seronegative partners, OR = 3.29, 95% CI = 2.16–5.01, were more likely to be aware than were partners of unknown serostatus. The interaction between partner type and serostatus was not significant. HIV-Negative and Unknown Serostatus Partner Awareness and HIV Transmission Risk Partner Awareness and Demographic Variables Because of the potential risk of HIV transmission to uninfected partners, we conducted additional analyses that addressed awareness among non-primary partners whom the participant believed were HIV-seronegative or whose HIV serostatus was not known by the participant. Most (85%; n = 309) participants with non-primary partners had sex with an HIV-seronegative or unknown serostatus non-primary partner. Of these 309 participants, 282 participants with complete data on all psychological, behavioral, and medical variables were included in analyses examining differences between partner awareness groups (See Table II). There were no differences between participants with and without complete data on demographic, psychological, behavioral, or medical variables. Among the 282 participants with complete data, 127 (45%) had nonprimary HIV-seronegative or unknown serostatus partners who were consistently unaware, 83 (29%) had non-primary HIV-seronegative or unknown serostatus partners who were inconsistently aware, and 72 (26%) had non-primary HIV-seronegative

or unknown serostatus partners who were consistently aware of the participant’s HIV serostatus. Men whose partners were inconsistently aware reported that 40% (M = 0.40; SD = 0.27) of these partners on average were aware (range 2–98%). The three partner awareness groups did not differ in age, income, education, employment, racial/ethnic background, recruitment city, or religion, years since testing HIV-positive, or in social desirability response bias. Participants with partners who were unaware were significantly less likely to have been diagnosed with AIDS than those with partners who were consistently aware or inconsistently aware. Partner Awareness and Psychological Variables Among men with one or more non-primary partners, partner awareness groups differed significantly in disclosure self-efficacy and disclosure intention, with those with partners who were consistently aware scoring higher than those with partners who were inconsistently aware, and those with partners who were inconsistently aware scoring higher than those with partners who were unaware. Groups also differed in comfort talking about HIV with non-primary partners, with those with partners who were consistently or inconsistently aware reporting greater comfort than those with partners who were unaware. Partner awareness groups did not differ in preferences of disclosing nonverbally versus verbally, comfort talking about HIV with close friends or with physicians or nurses, gay self-identification, or internalized homophobia. Partner Awareness and Sexual Practices Differences among partner awareness groups in sexual behavior related to HIV transmission risk and in use of venues to meet sexual partners were also

Partner Awareness of HIV Serostatus

161

Table II. Psychological, Behavioral, and Medical Differences Among Partner Awareness Groups of Men With Non-Primary Partners of HIV-Negative or Unknown Serostatus (n = 282‡ ) Partner awareness group

Psychological variables Disclosure self-efficacy (SD) Disclosure intention (SD) Prefer to disclose nonverbally Comfort talking about HIV serostatus With close friends (SD) With doctors or nurses (SD) With non-primary partners (SD) Internalized homophobia (SD) Gay self-identification Behaviors in the past 3 months Number of male partners Number of negative or unknown serostatus male partners (SD) All partners were one-night-stands Attended a sex venue to meet sex partners HIV transmission behavior Unprotected receptive anal Unprotected insertive anal Unprotected insertive oral Medical variables AIDS diagnosis Years since testing HIV+

F or χ2

Unaware (n = 127)

Inconsistently aware (n = 83)

Consistently aware (n = 72)

39.71∗∗∗ 45.46∗∗∗ 4.57

2.98 (0.99)a 2.43 (0.94)a 18%

3.66 (0.87)b 2.89 (1.06)b 14%

4.18 (0.90)c 3.90 (1.20)c 7%

0.58 0.01 22.02∗∗∗ 1.01 1.47

3.94 (1.23) 4.38 (1.02) 2.45 (1.15)a 2.04 (0.79) 94%

4.11 (1.18) 4.39 (1.06) 3.14 (1.22)b 1.88 (0.71) 92%

4.08 (1.35) 4.35 (1.16) 3.60 (1.33)b 2.01 (0.90) 89%

2.51∗∗∗

15.39 (23.28)a

21.42 (24.04)b

14.82 (21.44)c

4.39∗∗ 10.78∗∗ 13.81∗∗

12.99 (19.46)a 46%a 83%a

19.42 (22.77)b 29%b 85%a

10.43 (16.79)a 24%b 65%b

8.40 1.53 18.44∗

32%a 22% 75%a

49%b 32% 84%a

28%a 25% 61%b

10.49∗∗ 0.64

35%a 6.56 (3.95)

50%b 7.18 (3.96)

58%b 6.63 (3.95)

Note. df = 2 for each omnibus test; SD = standard deviation; numbers with different superscripts in a given row differ significantly at p < 0.05. F-test used for numeric data and χ2 test used for percentage data. ‡ Total n = 282 due to missing data. ∗ p < 0.05; ∗∗ p < 0.01; ∗∗∗ p < 0.001.

explored (see Table II). Those with partners who were inconsistently aware had a greater number of male partners and number of HIV-negative or unknown serostatus partners than those in the other two groups. Men with partners who were unaware had a higher number of male partners than those with partners who were consistently aware. Men with partners who were unaware were more likely to report all of their non-primary partners were one-night stands than those with partners who were consistently or inconsistently aware. Regarding places that men went to meet sex partners, those with partners who were unaware and inconsistently aware were more likely to have met a sex partner at sex venues than those with partners who were consistently aware.

Multivariate Analyses Predicting Partner Awareness Number of male sex partners and number of HIV-negative and unknown sex partners were cor-

related at r = .87. To avoid multicollinearity, only the variable of most relevance to HIV transmission, number of HIV-negative and unknown sex partners, was included in the multivariate model. Variables found to be associated with partner awareness group were entered in as a group into a multinomial logistic regression with partner awareness group as the dependent variable (see Table III). The model was significant, χ2 (14, N = 282) = 133.13, p < .01. Compared to men with partners who were unaware, men with partners who were consistently aware continued to have higher disclosure self-efficacy, have greater disclosure intention, have greater comfort talking about their HIV status, and be more likely to have an AIDS diagnosis. Compared to men with partners who were inconsistently aware, men with partners who were unaware had lower disclosure selfefficacy, slightly fewer non-primary HIV-negative or unknown serostatus partners (OR = 0.98) were more likely to report that all partners were one night stands and were less likely to have an AIDS diagnosis. Further, men with partners who were consistently

162

´ Hart, Wolitski, Purcell, Parsons, Gomez, and the Seropositive Urban Men’s Study Team

Table III. Multinomial Logistic Regression Predicting Differences Between Partner Awareness Groups of Men With Non-Primary Partners of HIV-Negative or Unknown Serostatus (n = 282) OR (95% CI) Predictors Psychological variables Disclosure self-efficacy Disclosure intention Comfort talking about HIV to non-primary partners Sexual behaviors in the past 3 months Number of HIV-negative or unknown serostatus non-primary partners Attending sex venues to meet partners All partners one night stands AIDS diagnosis

Consistently aware versus unaware

Unaware versus inconsistently aware

Consistently aware versus inconsistently aware

2.02 (1.22–3.34)∗∗ 2.33 (1.58–3.44)∗∗∗ 1.53 (1.06–2.19)∗

0.59 (0.41–0.86)∗∗ 0.75 (0.55–1.03) 0.77 (0.58–1.01)

1.46 (0.92–2.34) 1.89 (1.34–2.67)∗ 1.14 (0.84–1.54)

1.01 (0.99–1.04)

0.98 (0.96–0.99)∗∗

0.99 (0.97–1.01)

0.64 (0.25–1.63) 0.39 (0.17–1.01) 2.41 (1.06–5.49)∗

1.13 (0.48–2.66) 2.51 (1.28–4.92)∗∗ 0.50 (0.27–0.96)∗

0.55 (0.24–1.27) 0.94 (0.41–2.14) 1.24 (0.60–2.57)

Note. Table compares men with non-primary HIV-seronegative or unknown serostatus partners who were consistently aware, inconsistently aware and unaware the participant was HIV-seropositive. The last category in each column is the reference group. n = 282. ∗ p < 0.05; ∗∗ p < 0.01; ∗∗∗ p < 0.001.

aware had greater disclosure intention than those with partners who were inconsistently aware. There were no differences in number of HIV-seronegative or unknown serostatus partners between men with partners who were consistently aware and inconsistently aware. To examine the effect of preferences for verbal disclosure upon analyses predicting partner awareness, analyses were repeated selecting only those men who preferred to disclose verbally. An identical pattern of results emerged for both univariable and multivariate analyses.

Partner Awareness and HIV Transmission Behavior with Non-Primary Partners We examined differences in unprotected insertive and receptive anal intercourse and insertive oral intercourse with HIV-seronegative or unknown serostatus non-primary partners (see Table II). Men with partners who were inconsistently aware were more likely to engage in unprotected receptive anal intercourse than those whose partners were consistently aware or unaware. Although a similar pattern was observed for insertive anal intercourse, differences between groups were not significant. Regarding insertive oral intercourse, men with partners who were inconsistently aware and men with partners who were unaware were more likely to engage in unprotected insertive oral intercourse than those with partners who were consistently aware. Given that men with partners who were inconsistently aware ranged widely from 2 to 98% partner

awareness among HIV-seronegative or unknown serostatus partners, we conducted a sensitivity analysis by excluding outliers by 5% on each side. This resulted in the inconsistent partner awareness group including only those who disclosed 7 to 93% of the time. An identical pattern of results emerged when analyses were conducted excluding outliers from the inconsistent partner awareness group. Proportion of aware partners relative to total partners was also assessed as a potential predictor of sexual transmission risk. Proportion of aware partners was not associated with sexual transmission risk variables. Multivariate analyses examined whether partner awareness group would predict unprotected sexual behavior controlling for variables found to be associated with partner awareness group in the previous multinomial logistic regression. Regarding unprotected receptive anal intercourse, variables associated with partner awareness group were entered as a group in block 1 and this block failed to reach significance, χ2 (6, N = 282) = 12.22, n.s. Partner awareness group was entered in block 2 and was significant, χ2 (2, N = 282) = 9.42, p < 0.01. Compared to men with partners who were inconsistently aware, men with partners who were unaware continued to be less likely to have unprotected receptive anal intercourse, OR = 0.36, 95% CI = 0.18–0.70, but men with partners who were consistently aware were no longer less likely to have unprotected receptive anal intercourse, OR = 0.57, 95% CI = 0.27–1.19. An identical procedure was conducted examining unprotected insertive anal intercourse and insertive oral intercourse. Regarding insertive anal

Partner Awareness of HIV Serostatus intercourse, block 1 was significant, χ2 (6, N = 282) = 20.10, p < .01, with number of HIV-negative and unknown serostatus partners associated with increased likelihood of unprotected insertive anal intercourse, OR = 1.02, 95% CI = 1.01–1.04. Partner awareness was entered in block 2 and was significant, χ2 (2, N = 282) = 6.133, p < .05. Partner awareness group continued to be a significant predictor, with those with partners who were unaware being less likely to engage in this behavior than those with partners who were inconsistently aware, OR = 0.40, 95% CI = 0.19–0.84. Regarding insertive oral intercourse, block 1 was significant, χ2 (6, N = 282) = 40.07, p < .01, with disclosure self-efficacy, OR = 1.77, 95% CI = 1.22– 2.56, disclosure intention, OR = 0.73, 95% CI = 0.54–0.99, number of HIV-negative or unknown serostatus partners, OR = 7.88, 95% CI = 3.31– 18.81, and all partners being one-night stands, OR = 0.46, 95% CI = 0.23–0.91, predicting unprotected insertive oral intercourse. AIDS status and comfort talking about HIV failed to predict unprotected insertive oral intercourse. Partner awareness was entered in block 2 and was not significant, χ2 (2, N = 282) = 1.24, p < .05, with partner awareness no longer predicting unprotected insertive oral intercourse in the final model for unprotected insertive oral intercourse. Given the strong association of number of HIV-negative or unknown sersostatus partners with insertive oral intercourse, post-hoc analyses were conducted excluding this variable as a predictor. Excluding this variable, when partner awareness was entered in block 2 this block became significant, χ2 (2, N = 282) = 7.68, p < .05. Partner awareness predicted unprotected insertive oral intercourse, as men with partners who were consistently aware, OR = 0.38, 95% CI = 0.16–0.92, and men with partners who were unaware, OR = 0.37, 95% CI = 0.15–0.83, being less likely to engage in this behavior than those with partners who were inconsistently aware. To examine the effects of reported preference for verbal disclosure, analyses were repeated selecting only those men who preferred to disclose verbally. An identical pattern of results emerged for both univariate and multivariate analyses.

DISCUSSION Findings from the current study indicate that although serostatus disclosure is common for most

163 seropositive MSM, only 68% of HIV-negative or unknown serostatus primary partners in the past 3 months were aware the participant was HIVseropositive before having sex with him for the first time. Regarding non-primary partners, only 26% of HIV-seropositive MSM reported all non-primary HIV-negative or unknown serostatus partners were aware. There are important differences between men with non-primary partners who are HIVseronegative or of unknown serostatus partners who are all aware, some of whom are aware and none of whom are aware the participant was HIVseropositive. These groups significantly differed in disclosure self-efficacy, disclosure intention, comfort in talking about one’s HIV status, having all of one’s non-primary partners being one-night stands, AIDS diagnosis, and, most importantly, risk of HIV transmission. The current study presents a novel method for examining HIV serostatus partner awareness. Findings suggest that assessment of disclosure as a binary variable (e.g., who disclosed to all partners or not) may be insufficient, as more than a quarter of the participants reported their non-primary HIVnegative or unknown serostatus partners were inconsistently aware. as more of a quarter of those HIV+ participants reported their non-primary HIV- or unknown serostatus partners were inconsistently ware of their HIV+ serostatus. The relation between partner awareness and unprotected intercourse also does not appear to be linear, as inconsistency in partner awareness was related to increased sexual transmission risk but proportion of aware partners was not. Participants with partners who were inconsistently aware were more likely to report unprotected sexual behavior than those with partners who were consistently aware or unaware. Specifically, men with partners who were inconsistently aware were more likely to have had unprotected receptive anal and insertive anal intercourse. For insertive anal intercourse, partner awareness was significantly associated with this behavior only after controlling for correlates of partner awareness. Partners may be less likely to become aware when an HIV-seropositive MSM is engaging in sexual practices that carry a lower risk of HIV transmission than insertive anal intercourse. Inconsistency in partner awareness may explain discrepant findings in the literature, with some suggesting a link between disclosure and unprotected sex (De Rosa and Marks, 1998; Niccolai et al., 1999; Stein et al., 1998) and some failing to find an

164

´ Hart, Wolitski, Purcell, Parsons, Gomez, and the Seropositive Urban Men’s Study Team

association (Kalichman and Nachimson, 1999; Marks and Crepaz, 2001; Wolitski et al., 1998). That is, among studies examining disclosure in one sexual event (Marks and Crepaz, 2001) or to one sexual partner (Niccolai et al., 1999), men with partners who were inconsistently aware could have a partner who was made aware in one situation, thereby entering the “disclosed” group in one study, and not make their partners aware in another situation, thereby entering the non-disclosed group in a 2nd study. Inconsistent partner awareness may also explain null findings in studies comparing men who disclosed to all partners versus those who did not in a certain time period (De Rosa and Marks, 1998; Kalichman and Nachimson, 1999; Stein et al., 1998; Wolitski et al., 1998) because the higher rates of unprotected intercourse found among men who disclosed to some partners would be combined with the relatively lower rates of unprotected intercourse among men who disclosed to no partners. However, more research is needed to explicitly examine how inconsistent verbal disclosure is associated with risky sexual behavior, as the present study assessed partner awareness and not disclosure behavior. Assessment of partner awareness as a binary variable also may be insufficient because nonawareness of HIV-serostatus is not automatically associated with increased risk behavior. In fact, those with partners who were consistently aware and unaware did not differ in unprotected insertive or receptive anal intercourse. Further, those with partners who were unaware and those with partners who were inconsistently aware did not differ in unprotected insertive oral intercourse. Men with partners who were inconsistently aware occupied an intermediate position between the unaware and consistently aware groups on psychological variables relevant to disclosure. Men with partners who were inconsistently aware reported significantly lower disclosure self-efficacy and disclosure intention than those with consistently aware partners, but significantly higher disclosure self-efficacy and disclosure intention than men with unaware partners. These findings, combined with the increased sexual transmission behavior among the inconsistent partner awareness group even when controlling for psychological and behavioral correlates, suggest that increased disclosure-facilitating cognition is not necessarily associated with decreased sexual risk. Alternatively, other psychological variables not assessed in the present study may be more strongly implicated in the increased risk among those with partners who

were inconsistently aware (e.g., impulsivity; Wulfert et al., 1999). Previous research has found that HIVseropositive MSM with anonymous partners had lower disclosure self-efficacy relative to those with other types of casual or steady partners (Semple et al., 2000). Thus, many MSM who do not believe they will disclose may also choose to have anonymous partners, one-night stands, or meet partners in sex venues because they do not believe it is necessary to make partners aware in these contexts (Stirratt, 2005). Alternatively, HIV-seropositive men who have had sexual intercourse without their partners becoming aware of their serostatus may be less likely to continue sexual relationships (Schnell et al., 1992) because of fear associated with negative reactions upon disclosure (Gielen et al., 1997; Hays et al., 1993; Mansergh et al., 1995; Serovich, 2001; Stein et al., 1998). However, the present study examined overall patterns of partner awareness across a 3-month time period and cannot assess cognitions and emotions associated with one’s partner becoming aware or partner awareness versus lack of awareness in individual sexual situations. In sum, most HIV-seropositive men report their primary partners are aware and most men report at least some non-primary partners are aware the participant was HIV-seropositive before first having sex. Men appeared to differ in their overall pattern of partner awareness regarding HIV-negative or unknown serostatus non-primary partners. Those with partners who were consistently unaware had lower disclosure self-efficacy, disclosure intention, were more likely to report all partners were one-night stands, and were less likely to have an AIDS diagnosis than those with partners who were consistently and inconsistently aware. A much different pattern emerged regarding partner awareness and HIV transmission risk, suggesting partner awareness is related to safer sexual behavior in a non-linear fashion. Men with non-primary HIV-negative or unknown partners who were inconsistently aware had higher rates of unprotected receptive anal intercourse than those with partners who were consistently aware or consistently unaware and higher rates of insertive oral intercourse than those with partners who were consistently aware. However, there were no differences between those in the consistent partner awareness and partner unawareness groups. Assessments of partner awareness or overt disclosure as a linear variable (e.g., number of disclosure occasions, proportion of aware partners) or binary

Partner Awareness of HIV Serostatus variable (e.g., disclosure to recent partner or not) therefore may have failed to account for differences in HIV transmission behavior related to partner awareness. Further research is recommended to more fully explore possible relations between partner awareness inconsistency and HIV transmission risk. ACKNOWLEDGMENTS This research was conducted as part of the Seropositive Urban Men’s Study (SUMS). It was supported by the Centers for Disease Control and Prevention through cooperative agreements with New Jersey City University (U62/CCU213605), Rutgers University (U62/CCU2133607), and University of California, San Francisco (U62/CCU913557). The authors acknowledge the following SUMS collaborators for their contributions: Michael Stirratt, Robert Remien, Jeffrey Parsons, Ann O’Leary, Colleen Hoff, Robert Hays, James Carey, and Timothy Ambrose. The authors also gratefully acknowledge Gary Marks and Nicole Crepaz for comments on an earlier version of the manuscript. In addition, the authors thank the members of the SUMS community advisory boards in New York City and San Francisco for their guidance and thoughtful feedback. REFERENCES Crepaz, N., and Marks, G. (2003). Serostatus disclosure, sexual communication, and safer sex in HIV-positive men. AIDS Care, 15, 379–387. Crowne, D., and Marlowe, D. (1964). The approval motive. New York: Wiley. De Rosa, C. J., and Marks, G. (1998). Preventive counseling of HIV-positive men and self-disclosure of serostatus to sex partners: New opportunities for prevention. Health Psychology, 17, 224–231. DiFranceisco, W., McAuliffe, T. L., and Sikkema, K. J. (1998). Influences of survey instrument format and social desirability on the reliability of self-reported high-risk sexual behavior. AIDS and Behavior, 2, 329–337. Gielen, A. C., O’Campo, P., Faden, R. R., and Eke, A. (1997). Women’s disclosure of HIV status: Experiences of mistreatment and violence in an urban setting. Women and Health, 25, 19–31. Hays, R. B., McKusick, L., Pollack, L., Hilliard, R., Hoff, C., and Coates, T. J. (1993). Disclosing HIV seropositivity to significant others. AIDS, 7, 425–431. Kalichman, S. C., and Nachimson, D. (1999). Self-efficacy and disclosure of HIV-positive serostatus to sex partners. Health Psychology, 18, 281–287. Kirk, R. E. (1968). Experimental design: Procedures for the basic sciences. Belmont, CA: Brooks-Cole. Mansergh, G., Marks, G., and Simoni, J. M. (1995). Self-disclosure of HIV infection among men who vary in time since

165 seropositive diagnosis and symptomatic status. AIDS, 9, 639– 644. Marks, G., Bundek, N. I., Richardson, J. L., Ruiz, M. S., Maldonado, N., and Mason, H. R. C. (1992). Self-disclosure of HIV infection: Preliminary results from a sample of Hispanic men. Health Psychology, 11, 300–306. Marks, G., and Crepaz, N. (2001). HIV-positive men’s sexual practices in the context of self-disclosure of HIV status. Journal of Acquired Immune Deficiency Syndromes, 27, 79– 85. Niccolai, L. M., Dorst, D., Myers, L., and Kissinger, P. J. (1999). Disclosure of HIV status to sexual partners: Predictors and temporal patterns. Sexually Transmitted Diseases, 26, 281– 185. Ostrow, D. G., Joseph, J. G., Kessler, R., Soucy, J., Tal, M., Eller, M., Chmiel, J., and Phair, J. P. (1989). Disclosure of HIV antibody status: Behavioral and mental health correlates. AIDS Education and Prevention, 1, 1–11. Perry, S. W., Card, C. A. L., Mofatt, M., Jr., Ashman, T., Fishman, B., and Jacobsberg, L. B. (1994). Selfdisclosure of HIV infection to sexual partners after repeated counseling. AIDS Education and Prevention, 6, 403– 411. Purcell, D. W., Parsons, J. T., Halkitis, P. N., Mizuno, Y., and Woods, W. J. (2001). Substance use and sexual transmission risk behavior of HIV-positive men who have sex with men. Journal of Substance Abuse, 13, 185–200. Reynolds, W. M. (1992). Development of reliable and valid short forms of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical Psychology, 38, 119–125. Ross, M. W., and Rosser, B. R. S. (1996). Measurement and correlates of internalized homophobia: A factor analytic study. Journal of Clinical Psychology, 52, 15– 21. Schnell, D. J., Higgins, D. L., Wilson, R. M., Goldbaum, G., Cohn, D. L., and Wolitski, R. J. (1992). Men’s disclosure of HIV test results to male primary sex partners. American Journal of Public Health, 82, 1675–1676. Semple, S. J., Patterson, T. L., and Grant, I. (2000). Partner type and sexual risk behavior among HIV positive gay and bisexual men: Social cognitive correlates. AIDS Education and Prevention, 12, 340–356. Serovich, J. M. (2001). A test of two disclosure theories. AIDS Education and Prevention, 13, 355–364. Simoni, J. M., Mason, H. R. C., and Marks, G. (1997). Disclosing HIV status and sexual orientation to employers. AIDS Care, 9, 589–599. Stein, M. D., Freedberg, K. A., Sullivan, L. M., Savetsky, J., Levenson, S. M., Hingson, R., and Samet, J. H. (1998). Sexual ethics: Disclosure of HIV-positive status to partners. Archives of Internal Medicine, 158, 253–257. Stein, M. D., and Samet, J. H. (1999). Disclosure of HIV status. AIDS Patient Care and STDs, 13, 265–267. Stirratt, M. J. (2005). HIV serostatus disclosure strategies and sex´ ual practices: Intimate connections. In P. Halkitis, C. Gomez, and R. J. Wolitski (Eds.) Positive Living: The Sexual Lives of Seropositive Gay and Bisexual Men. Washington, DC: American Psychological Association, pp. 101–120. Suarez, T., and Miller, J. (2001). Negotiating risks in context: A perspective on unprotected anal intercourse and barebacking among men who have sex with men. Archives of Sexual Behavior, 30, 287–300. Van de Ven, P., Kippax, S., Crawford, J., Rawstorne, P., Prestage, G., Grulich, A., and Murphy, D. (2002). In a minority of gay men, sexual risk practice indicates strategic positioning for perceived risk reduction rather than unbridled sex, AIDS Care, 14, 471–480. Winer, B. J. (1971). Statistical Principles in Experimental Design. New York: McGraw-Hill.

166

´ Hart, Wolitski, Purcell, Parsons, Gomez, and the Seropositive Urban Men’s Study Team

Wolitski, R. J., Halkitis, P. N., Parsons, J. T., and Gomez, C. A. (2001). Awareness and use of untested barrier methods by HIV-seropositive gay and bisexual men. AIDS Education and Prevention, 13, 291–301. Wolitski, R. J., Rietmeijer, C. A. M., Goldbaum, G. M., and Wilson, R. M. (1998). HIV serostatus disclosure among gay and bisexual men in four American cities: General patterns and relation to sexual practices. AIDS Care, 10, 599–610.

Wulfert, E., Safren, S. A., Brown, I., and Choi, W. K. (1999). Cognitive, behavioral, and personality correlates of HIV-positive persons’ unsafe sexual behavior. Journal of Applied Social Psychology, 29, 223–244. Zea, C. M., Reisen, C. A., Poppen, P. J., and Diaz, R. M. (2003). Asking and telling: Communication about HIV status among Latino HIV-positive gay men. AIDS and Behavior, 7, 143– 152.