Abstract
Purpose
There is growing interest in network-based interventions to reduce HIV sexual risk behavior among both homeless youth and men who have sex with men (MSM). The goal of this study is to better understand the social network and individual correlates of sexual risk behavior among homeless MSM youth in order to inform these HIV prevention efforts.
Methods
A multi-stage sampling design was used to recruit a probability sample of 121 homeless MSM youth (ages 16–24) from shelters, drop-in centers, and street venues in Los Angeles County. Face-to-face interviews were conducted. Due to the different distributions of the three outcome variables, three distinct regression models were needed: ordinal logistic regression for unprotected sex; zero-truncated Poisson regression for number of sex partners; and logistic regression for any sex trade.
Results
Homeless MSM youth were less likely to engage in unprotected sex and had fewer sex partners if their networks included platonic ties to peers who regularly attended school, and had fewer sex partners if most of their network members were not heavy drinkers. Most other aspects of network composition were unrelated to sexual risk behavior. Individual predictors of sexual risk behavior included being older, Hispanic, lower education, depressive symptoms, less positive condom attitudes, and sleeping outdoors because of nowhere else to stay.
Conclusions
HIV prevention programs for homeless MSM youth may warrant a multi-pronged approach that helps these youth strengthen their ties to prosocial peers, develop more positive condom attitudes, and access needed mental health and housing services.
Keywords: homeless, youth, MSM, HIV, condom use, multiple partners, sex trade
INTRODUCTION
HIV/AIDS poses a significant public health threat among homeless youth given their high rates of risky sexual and drug-using behaviors. [1–4] Homeless young men who have sex with men (YMSM) deserve particular attention, as living on the streets or being marginally housed may further increase their vulnerability to HIV/AIDS. [5, 6] Studies generally find greater sexual risk behavior among homeless GLB youth compared to their heterosexual counterparts. [7, 8] Further, homeless gay and lesbian youth are more likely to report being diagnosed and treated for HIV compared to homeless bisexual and heterosexual youth. [9] Better understanding the social contextual and individual determinants of sexual risk behavior among homeless YMSM is critical to HIV prevention efforts with this vulnerable population. Social network analysis can be a useful tool for understanding the social contextual factors relevant to engagement in risk behavior, and finding that sexual risk behavior among homeless YMSM is influenced by characteristics of their social networks may lead to innovative and potentially more effective intervention approaches for this population.
Only two studies have focused on the egocentric (personal) networks of homeless youth, which are comprised of the people that the youth knows, to examine associations between a wide range of network characteristics and their engagement in sexual risk behavior. One study was based on up to 7 individuals [10] and the other based on up to 2 individuals [11] with whom the youth had frequent contact. Neither study focused specifically on homeless YMSM. These studies suggest that homeless youth are more likely to engage in sexual risk behavior if their networks are less emotionally close, but provide more support; do not include family members; and contain individuals who either engage in risk behavior or encourage the youth to do so. However, these findings should be interpreted with the following caveats. Relatively few significant associations emerged from multivariate analyses relative to the number of examined associations, and results from the one study comparing different sexual risk behaviors suggests that social network influences may be largely restricted to multiple sexual partnerships. In addition, it is possible that these apparent network effects should be attributed more specifically to the influence of sex partners. Sex partners comprise a large proportion of the networks of sexually active homeless youth and are a primary source of both risk and support; [12] thus, egocentric network studies that elicit a very small number of close ties may be focusing more on the influence of sex partners than other types of network members. The present study is designed to address these issues in examining the social network correlates of HIV sexual risk behavior among homeless YMSM.
In terms of individual characteristics, existing literature points to several factors that are relevant to sexual risk behavior among homeless youth and thus warrant examination in this study. To the extent that these individual characteristics are also correlated with the composition of youth’s networks, including them is particularly important in that they may be responsible for any associations that are found between network composition and sexual risk behavior. For example, homeless substance using youth are both more likely to engage in risky sex [4, 13, 14] and to have networks that include substance users. [10, 12, 15] Thus, an apparent association between exposure to substance using network members and the youth’s engagement in risky sex may actually be due to the youth’s own substance use. Similarly, poor mental health, history of child abuse, and homelessness severity have been associated with sexual risk behavior among homeless youth [16–19] and network composition in other populations. [20, 21]
The present study addresses the following two research questions. First, when simultaneously considering both social network and individual factors, what are the correlates of sexual risk behavior among homeless MSM youth? In answering this question, we were interested in examining three distinct sexual risk behaviors (unprotected anal sex, number of sex partners, and sex trade) that may differ in their risk and protective factors, including the importance of social network influences. Second, if social network correlates of sexual risk behavior are identified, do they appear to be primarily due to the potential influence of sex partners, platonic network ties, or both? We use a more comprehensive approach in eliciting network information than previous studies of homeless youth in order to address this question.
METHODS
Participants
Youth were eligible if they: a) were between the ages of 13–24; b) were not currently living with a parent or guardian; c) were not getting most of their support for food and housing from family or a guardian; d) spent the previous night in a shelter, outdoor or public place, hotel or motel room rented with friends (because of no place else to go), or other place not intended as a domicile; and e) were English speaking. In this study we focus on 121 male youth who reported having sex with a male partner during the past 3 months (see Table 1). Written informed consent was obtained from participants prior to conducting a computer-assisted face-to-face structured interview. Interviews occurred between October 2008 and December 2009. The research protocol was approved by the institutional review board of RAND and a Certificate of Confidentiality was obtained from the USDHHS.
TABLE 1.
Descriptive Statistics for Main Study Variables (Weighted)
| Variables | Min | Mean (SD) or % |
Median | Max |
|---|---|---|---|---|
| Age in years | 16 | 21.02 (0.22) | 21 | 24 |
| Race/ethnicity | ||||
| African American | 33% | |||
| Hispanic | 21% | |||
| White | 17% | |||
| Other or multiracial | 29% | |||
| High school diploma or GED | 71% | |||
| Any history of child abuse | 72% | |||
| Depressive symptoms | 0 | 1.05 (0.08) | 1 | 3 |
| Positive condom beliefs | 1 | 2.99 (0.08) | 3 | 4 |
| Homelessness severity | ||||
| Any sleeping outside in past month | 60% | |||
| # of years homeless | 0 | 5.37 (0.40) | 5 | 18 |
| Substance use in past month | ||||
| # of days heavy drinking | 0 | 2.24 (0.67) | 0 | 30 |
| Any hard drug use | 38% | |||
| Location where sampled | ||||
| Shelter | 30% | |||
| Drop-in center or street venue | 70% | |||
| Network composition | ||||
| # of relatives | 0 | 4 (0.40) | 3 | 17 |
| Any adults in positions of responsibility | 29% | |||
| ≥ 50% drinks heavily | 32% | |||
| ≥ 50% uses drugs | 42% | |||
| # who engage in risky sex | 0 | 5 (0.62) | 3 | 20 |
| # who attend school regularly | ||||
| 0 | 16% | |||
| 1–2 | 32% | |||
| 3 or more | 52% | |||
| # who provide emotional support | 0 | 8 (0.67) | 5 | 20 |
| # who provide tangible support | 0 | 6 (0.72) | 4 | 20 |
| Sexual risk behavior in past 3 months | ||||
| Unprotected anal sex | ||||
| Never had unprotected sex | 50% | |||
| Sometimes had unprotected sex | 27% | |||
| Always had unprotected sex | 23% | |||
| # of sex partners, past 3 months | 1 | 6 (0.69) | 3 | 200 |
| Any trading sex | 27% |
Sampling Design
We recruited a probability sample of homeless youth from shelters, drop-in centers, and street venues in Los Angeles County. A multi-stage design was implemented in which we first developed a list of sites used by homeless youth and then drew a probability sample of homeless youth from these sites. Design details are reported elsewhere [22]. The main data collection for this study involved conducting 419 interviews with homeless youth sampled from 22 service sites and 19 street sites (78% eligibility rate; 97% response rate), which resulted in completed interviews with 45 YMSM. Near the end of the field period we conducted a supplemental data collection to accrue additional cases of YMSM, focusing our efforts on seven of the original 41 sites that yielded the highest recruitment of homeless YMSM and adding three new street sites where homeless YMSM in particular were known to congregate. The supplemental data collection resulted in completed interviews with 76 homeless YMSM [36% eligibility rate (lower than the main sample due to additional eligibility criteria of having a male partner in the past 3 months); 92% response rate), resulting in the final analytic sample of 121 YMSM.
Measures
Sexual risk behavior
Three behaviors were assessed with reference to the past 3 months: (a) unprotected sex (3-level: never had unprotected sex, sometimes had unprotected sex, or always had unprotected sex); (b) number of male or female sex partners; and (c) any sex trade (defined as any of their partners giving them something to have sex such as money, a place to stay, or alcohol or drugs).
Personal network characteristics
Participants provided the first names of 20 individuals over the age of 13 with whom they had mutual recognition (“people that you know and who know you”) and contact sometime during the past 3 months (face-to-face or by phone, mail, or e-mail). We used a general name generator (rather than asking youth to name particular types of people, such as close friends) to allow for identification of a greater diversity of network members (alters). A series of questions about each alter determined how the youth met the alter (if not a relative) and the perceived behaviors of the alter. Network composition measures include the number of alters who: a) were perceived by the youth as engaging in risky sex during the past 3 months; b) were relatives; c) provided emotion support; and d) provided tangible support. Several other network variables were recoded as categorical due to their distributions: a) ≥ 50% of network members used alcohol to the point of being drunk in the past 3 months; b) ≥ 50% of network members used illicit drugs in the past 3 months; c) any (non-relative) adult in a position of responsibility (e.g., service provider, teacher, boss, clergy); and d) number of network members who regularly attend school (1 or 2 versus 0, 3 or more versus 0).
Background and individual characteristics
Variables included age, race/ethnicity, education (had a high school diploma/GED versus not), history of child abuse while living at home, depressive symptoms, condom beliefs, severity of homelessness, substance use, and recruitment site (shelter versus drop-in center or street). Depressive symptoms were assessed with a 4-item version of the CESD [23, 24] (α=0.82) and we averaged ratings across these items. Positive condom beliefs were assessed with a 3-item scale asking whether condoms interfere with the enjoyment of sex, condoms can be used without ruining the mood, and you can stop before sex to use a condom [25, 26] (α=.56). Homelessness severity included whether the youth spent the night outdoors during the past 30 days because they had nowhere else to stay (1 = yes, 0 = no), and number of years since the youth left home for the first time. Past month substance use was measured in terms of heavy drinking (number of days consumed 5+ drinks in a row) and any hard drug use (which, in most cases, was methamphetamine or ecstasy).
Analyses
We first examined bivariate associations between the independent variables and the three sexual behavior outcome variables. For each outcome, independent variables associated at p < .10 were retained in a multivariate model. Due to the different distributions of the outcome variables, three distinct regression models were used: ordinal logistic regression for unprotected sex; zero-truncated Poisson regression for number of partners; and logistic regression for any sex trade. The first set of bivariate models included network measures that were calculated based on all 20 named network members. We then ran two additional sets of bivariate models using network measures based on: (1) network members who were recent sex partners; and (2) network members who were platonic ties. Given their distributions, the newly derived variables based on recent sex partners only were dichotomized as any partner versus no partners having the particular characteristic.
RESULTS
Descriptive Information on Sexual Behavior
As shown in Table 1, respondents reported an average of 6 sex partners during the past 3 months. Most youth (79.6%) included at least one recent sex partner in their list of 20 network members; on average, 54% of recent sex partners were included among these 20 network members. Half of respondents reported never having unprotected sex during the past 3 months, with the remainder almost evenly split between those who sometimes and always had unprotected sex. On average, youth reported engaging in unprotected sex 35% of the time they had anal sex. Twenty-seven percent of YMSM reported trading sex in the past 3 months.
Bivariate Associations With Sexual Risk Behavior
As shown in Table 2, unprotected anal sex was more likely among YMSM who were older, were Hispanic (versus White), had less education, had a history of child abuse, and held less positive attitudes about condoms. Unprotected sex was more likely among YMSM with more relatives in their network, but less likely among those with ties to network members who attended school regularly. In the case of number of sex partners, YMSM reported more sex partners if they were had a longer history of homelessness, were hard drug users, and had been sampled from a shelter (marginal). They also reported more sex partners if their personal networks included more heavy drinkers, individuals who engaged in risky sex, and tangible support providers. Fewer sex partners were reported by YMSM with 3 or more network members who were engaged in school and among African Americans (marginal). Sex trade was more likely among YMSM with greater depressive symptoms and who slept outside because of nowhere else to stay, as well as marginally more likely among those with a longer history of homelessness and who engaged in less frequent heavy drinking.
TABLE 2.
Bivariate Associations of Individual and Network Characteristics With Sexual Risk Behavior in Past 3 Months
| Predictor Variable | Unprotected Anal Sex |
Number of Sex Partners |
Any Trading Sex |
|||
|---|---|---|---|---|---|---|
| OR | (95% CI) | IRR | (95% CI) | OR | (95% CI) | |
| Individual characteristics | ||||||
| Age in years | 1.21 | (1.00, 1.46) | 1.04 | (0.90, 1.20) | 1.02 | (0.80, 1.32) |
| Black (versus white) | 1.11 | (0.37, 3.32) | 0.51 | (0.24, 1.10)# | 0.77 | (0.20, 2.88) |
| Hispanic (versus white) | 5.64 | (1.51, 21.10) | 0.84 | (0.35, 1.20) | 0.44 | (0.10, 2.00) |
| Other (versus white) | 0.98 | (0.28, 3.48) | 0.60 | (0.25, 1.48) | 0.33 | (0.07, 1.51) |
| High school diploma/GED | 0.32 | (0.13, 0.76) | 0.62 | (0.34, 1.12) | 0.63 | (0.22, 1.78) |
| Any history of child abuse | 2.37 | (1.02, 5.54) | 1.26 | (0.81, 1.97) | 1.39 | (0.50, 3.88) |
| Depressive symptoms | 1.07 | (0.66, 1.73) | 1.30 | (0.79, 2.16) | 2.19 | (1.23, 3.92) |
| Positive condom beliefs | 0.53 | (0.30, 0.93) | 0.87 | (0.49, 1.55) | 0.64 | (0.35, 1.15) |
| Any sleeping outside in past month | 0.96 | (0.40, 2.29) | 1.14 | (0.64, 2.02) | 3.92 | (1.31, 11.78) |
| # of years homeless | 1.04 | (0.89, 1.22) | 1.08 | (1.01, 1.15) | 1.13 | (0.99, 1.30)# |
| # of days heavy drinking | 1.03 | (0.94, 1.13) | 1.00 | (0.96, 1.03) | 0.93 | (0.85, 1.01)# |
| Any hard drug use | 1.01 | (0.43, 2.36) | 1.93 | (1.11, 3.38) | 1.07 | (0.41, 2.78) |
| Sampled from a shelter | 0.63 | (0.27, 1.45) | 1.72 | (0.96, 3.10)# | 1.07 | (0.43, 2.70) |
| Network characteristics a | ||||||
| # of relatives | 1.13 | (1.01, 1.26) | 0.96 | (0.90, 1.03) | 1.02 | (0.88, 1.18) |
| Any adult in responsible position | 0.71 | (0.28, 1.80) | 0.69 | (0.39, 1.22) | 0.67 | (0.22, 2.03) |
| ≥ 50% drink heavily | 1.07 | (0.45, 2.53) | 2.33 | (1.41, 3.86) | 1.00 | (0.36, 2.79) |
| ≥ 50% use drugs | 0.95 | (0.42, 2.15) | 1.07 | (0.60, 1.89) | 0.81 | (0.31, 2.11) |
| # who engage in risky sex | 1.02 | (0.93, 1.12) | 1.08 | (1.04, 1.12) | 1.01 | (0.95, 1.08) |
| # who attend school regularly | ||||||
| 1 or 2 (versus 0) | 0.21 | (0.06, 0.81) | 0.68 | (0.34, 1.36) | 2.84 | (0.69, 11.63) |
| 3 or more (versus 0) | 0.26 | (0.07, 0.89) | 0.45 | (0.25, 0.81) | 2.31 | (0.61, 8.69) |
| # who provide emotional support | 1.00 | (0.94, 1.07) | 1.01 | (0.97, 1.06) | 1.01 | (0.93, 1.09) |
| # who provide tangible support | 1.01 | (0.96, 1.06) | 1.04 | (1.01, 1.08) | 1.02 | (0.95, 1.10) |
Note.
p < .10.
Network measures were calculated based on all 20 named network members.
To better understand the network associations found for unprotected sex and number of sexual partnerships, we ran two additional sets of bivariate models using network measures based on: (1) network members who were recent sex partners; and (2) platonic network members. Table 3 shows that homeless YMSM were less likely to have unprotected sex and had fewer sex partners if their networks included platonic ties to peers who regularly attended school. Homeless YMSM tended to have more sex partners if their network included more sex partners who provided tangible support, as well as more sex partners and platonic ties who were perceived to drink heavily and engage in risky sex. We used these more specific network measures in the multivariate models with the exception of heavy drinking, given the strong correlation between heavy drinking by sex partners and heavy drinking by platonic ties (r = .64).
TABLE 3.
Bivariate Associations of Sexual Risk Behavior with Characteristics of Network Members Who Are Platonic Vs. Recent Sex Partners
| Unprotected Anal Sex |
Number of Sex Partners |
|||
|---|---|---|---|---|
| Predictor Variable | OR | (95% CI) | IRR | (95% CI) |
| Network members who are platonic | ||||
| ≥ 50% drink heavily | 1.89 | (1.08, 3.33) | ||
| # who engage in risky sex | 1.05 | (0.99, 1.11)# | ||
| # who attend school regularly | ||||
| 1 or 2 (versus 0) | 0.27 | (0.07, 0.99) | 0.36 | (0.20, 0.68) |
| 3 or more (versus 0) | 0.43 | (0.13, 1.45) | 0.26 | (0.14, 0.47) |
| # who provide tangible support | 1.02 | (0.98, 1.06) | ||
| Network members who are recent sex partners | ||||
| Any drink heavily | 2.11 | (1.31, 3.40) | ||
| Any engage in risky sex | 3.05 | (1.96, 4.73) | ||
| Any attends school regularly | 0.40 | (0.13, 1.29) | 1.40 | (0.61, 3.25) |
| Any provides tangible support | 1.73 | (1.06, 2.82) | ||
Note.
p < .10.
Multivariate Associations With Sexual Risk Behavior
Table 4 shows results from the multivariate models that included all variables bivariately associated with a particular sexual risk behavior outcome at p < .10. Unprotected anal sex remained more likely among those who are older, of Hispanic ethnicity, and with a greater number of relatives in their network, and less likely among those with more education, who held more positive condom attitudes, and who had platonic ties to 1 or 2 individuals who regularly attended school. In addition, unprotected anal sex was marginally more likely among those with a history of child abuse, and less likely among who had platonic ties to 3 or more individuals who regularly attended school (p = .06 in both cases). Sex trade remained more likely among youth with depressive symptoms and who slept outside, but marginally less likely among those who engaged in more frequent heavy drinking. Length of homelessness was unrelated to sex trade in the multivariate model. Finally, homeless YMSM reported more sex partners if their networks were largely comprised of heavy drinkers and they had at least one sex partner who engaged in risky sex, but fewer sex partners if their networks included platonic ties to peers who regularly attended school. In addition, YMSM were marginally more likely to report a greater number of sex partners if they were hard drug users. Race, number of years homeless, the location where sampled, and the provision of tangible support by sex partners were not associated with number of sexual partnerships in the multivariate model.
TABLE 4.
Multivariate Associations of Individual and Network Characteristics With Sexual Risk Behavior in Past 3 Months
| Unprotected Anal Sex |
Number of Sex Partners |
Any Trading Sex |
||||
|---|---|---|---|---|---|---|
| Predictor Variable | OR | (95% CI) | IRR | (95% CI) | OR | (95% CI) |
| Individual characteristics | ||||||
| Age in years | 1.27 | (1.03, 1.58) | ||||
| Black (versus white) | 0.74 | (0.23, 2.37) | 0.57 | (0.26, 1.21) | ||
| Hispanic (versus white) | 4.73 | (1.11, 20.26) | 0.75 | (0.28, 2.00) | ||
| Other (versus white) | 0.73 | (0.18, 2.88) | 0.71 | (0.35, 1.46) | ||
| High school diploma/GED | 0.34 | (0.14, 0.86) | ||||
| Any history of child abuse | 2.49 | (0.96, 6.49)# | ||||
| Depressive symptoms | 1.94 | (1.02, 3.66) | ||||
| Positive condom beliefs | 0.49 | (0.26, 0.90) | ||||
| Any sleeping outside in past month | 3.87 | (1.23, 12.18) | ||||
| # of years homeless | 1.04 | (0.99, 1.10) | 1.10 | (0.95, 1.27) | ||
| # of days heavy drinking | 0.91 | (0.83, 1.00)# | ||||
| Any hard drug use | 1.48 | (0.98, 2.25)# | ||||
| Sampled from a shelter | 1.44 | (0.83, 2.64) | ||||
| Network characteristics | ||||||
| # of relatives | 1.17 | (1.03, 1.33) | ||||
| ≥ 50% of all network members drink heavily | 1.53 | (1.01, 2.31) | ||||
| Any recent sex partner engages in risky sex | 1.71 | (1.14, 2.56) | ||||
| # platonic ties who engage in risky sex | 0.98 | (0.93, 1.03) | ||||
| # platonic ties who attend school regularly | ||||||
| 1 or 2 (versus 0) | 0.16 | (0.05, 0.54) | 0.48 | (0.27, 0.86) | ||
| 3 or more (versus 0) | 0.34 | (0.11, 1.03)# | 0.36 | (0.23, 0.57) | ||
| Any recent sex partner provides tangible support | 1.31 | (0.85, 2.02) | ||||
Note.
p < .10.
DISCUSSION
Despite the diversity we found within the social networks of homeless YMSM, only the presence of certain types of network ties was associated with their sexual risk behavior. Homeless YMSM were less likely to engage in unprotected sex, and had fewer sex partners, if their networks included platonic peers who regularly attended school. They also tended to have fewer sex partners if most of their network members were not heavy drinkers. We found a similar association of having ties to peers who attend school and do not drink heavily with less substance use in our full sample of homeless youth [22], raising the possibility that these network members may serve as prosocial role models that have a general deterring effect on risk behavior. Although we cannot infer causality from our data, there is corroboration for this interpretation from a longitudinal study of newly homeless youth which found that unprotected sex was less likely if youth had more friends who attended school. [15] Given that the social networks of homeless youth tend to be largely comprised of individuals who are met on the street and engage in high-risk behavior, the presence of prosocial peers is often overlooked. Yet there is some evidence that homeless youth who keep in touch with low-risk, home-based peers, such as through social networking technology, are less likely to engage in risky sexual behavior. [27] Although the extent to which this is due to peer influence versus selection effects is unclear, it does raise the possibility that fostering these types of prosocial ties may be a promising approach to reducing sexual risk behavior.
Results from this study also indicate that unprotected sex is more likely among homeless YMSM who have more relatives in their network. This might seem unexpected in light of prior research indicating that the presence of a family member in the networks of homeless youth is associated with less sexual risk taking in general, [11] but the evidence has not been entirely consistent. [10] [17] There is not a clear explanation for why having relatives in the network is associated with an increased likelihood of unprotected sex. One possibility is that homeless youth who maintain ties to family are better able to form and maintain stable sexual partnerships, and those who are in stable sexual partnerships are less likely to use condoms. In our sample, however, it was not the case that homeless YMSM with primary sex partners were more likely to have relatives in their network compared to those with only casual partners (results not shown). Another possibility is that the association might be due to ethnicity if Hispanic MSM both have more family members in their network and are more likely to have unprotected sex. This idea was unsupported in our sample as well, given that the association remained significant after controlling for ethnicity. Further research is needed to better understand whether family ties might influence sexual risk behavior among homeless youth, or whether the presence of these ties is a proxy for another variable unmeasured in this study.
There is little evidence from this study that the sexual risk behaviors of homeless YMSM are influenced by the presence of adults in positions of responsibility, drug users, or individuals who provide emotional or tangible support. Although YMSM tended to report having a greater number of sex partners if their networks included more individuals who engaged in risky sex, additional analyses revealed that this association was mostly limited to network members who were recent sex partners. It is possible that social norms within the sexual networks of homeless YMSM encourage engagement in sexual risk behavior (at least in terms of having multiple sexual partnerships). However, the absence of any similar associations with condom use or sex trade behavior raises the possibility that youth with a greater number of sex partners are simply more likely to have a sex partner in their network who engages in risky sex.
In addition to better understanding the relevance of personal networks to HIV sexual risk behavior among homeless YMSM, our results highlight a number of important individual targets for HIV prevention programs developed for this population. For example, greater unprotected sex among Hispanics, reported as well in other studies of MSM, [28–32] suggests a need for cultural tailoring in interventions to encourage condom use among homeless YMSM. The marginally greater engagement in unprotected sex among homeless YMSM with a history of childhood abuse contributes to a growing literature suggesting the need for HIV prevention interventions to address the mental health issues of MSM with childhood abuse histories. [33, 34] The same argument for treatment need can be made for youth who are hard drug users or have a history of depressive symptoms, which may both contribute to continued homelessness and sexual risk behavior, as well as be exacerbated by these experiences. [35, 36] Finally, we found that homelessness severity was a risk factor for sex trade, with those in the most dire circumstance of sleeping on the streets being nearly four times as likely as other homeless YMSM to exchange sex for things they needed. The idea of providing stable housing as an HIV prevention strategy for homeless adults has gained momentum in recent years [22, 37, 38] and results from this study suggest that this approach merits attention in the case of homeless YMSM in the transition to young adulthood.
Strengths of this study include its rigorous sampling design, recruitment from both service and street sites (so that youth who do not seek services are represented), and collection of detailed personal network information. Nonetheless, the study is limited by its focus on Los Angeles County and results may not be generalizable to homeless YMSM living in other geographic areas. Further, the cross-sectional design is limiting in two respects: we do not know whether the composition of youth’s personal networks are associated with change in sexual risk behavior over time; and we do not know the extent to which associations between network characteristics and sexual risk behavior are due to selection effect (e.g., less risky youth may tend to seek out and form friendships with less risky peers). Limitations of our measures include the measure of unprotected anal sex not differentiating between different types of sex (e.g., insertive versus receptive sex, sex with casual versus primary partners), and the measure of positive condom attitudes having relatively low reliability. Finally, because the sample size was limited, the study was not powered to detect smaller effects. This might also explain the large confidence intervals obtained for some of the variables (e.g., being Hispanic).
Given the growing interest in network-based interventions to reduce sexual risk behavior among both MSM [39] and homeless youth, [27, 40] it is important to better understand the potential promise and limitations of this approach. Helping homeless YMSM strengthen their ties with prosocial peers, such as those who are actively engaged in school or refrain from heavy alcohol use, may provide positive role models that discourage their engagement in HIV sexual risk behavior. However, it is less clear that strengthening ties to relatives, adults in positions of responsibility, or support providers holds the same promise for this population. Our results suggest that a more promising prevention approach may involve strengthening their ties to prosocial peers, fostering positive attitudes about condoms, and providing supportive services that address their mental health and housing needs.
ACKNOWLEDGMENT
This research was supported by Grant R01DA020351 from the National Institute on Drug Abuse. We thank the youth who shared their experiences with us, the service agencies that collaborated in the study, and the RAND Survey Research Group for their assistance in data collection. Some of these data were previously presented at the 2011 meeting of the Society of Behavioral Medicine.
Footnotes
IMPLICATIONS AND CONTRIBUTIONS
HIV/AIDS poses a significant public health threat among homeless MSM youth, yet efficacious prevention programs for this population are lacking. Results suggest that helping youth strengthen ties to prosocial peers, develop positive condom attitudes, and access needed mental health and housing services may reduce sexual risk behavior in this population.
REFERENCES
- 1.De Rosa CJ, Montgomery SB, Hyde J, et al. HIV Risk Behavior and HIV Testing: A Comparison of Rates and Associated Factors among Homeless and Runaway Adolescents in Two Cities. AIDS Educ Prev. 2001;13:131–148. doi: 10.1521/aeap.13.2.131.19739. [DOI] [PubMed] [Google Scholar]
- 2.Kral AH, Molnar BE, Booth RE, et al. Prevalence of Sexual Risk Behaviour and Substance Use among Runaway and Homeless Adolescents in San Francisco, Denver and New York City. Int J STD AIDS. 1997;8:109–117. doi: 10.1258/0956462971919651. [DOI] [PubMed] [Google Scholar]
- 3.Tevendale HD, Lightfoot M, Slocum SL. Individual and Environmental Protective Factors for Risky Sexual Behavior among Homeless Youth: An Exploration of Gender Differences. AIDS Behav. 2009;13:154–164. doi: 10.1007/s10461-008-9395-z. [DOI] [PubMed] [Google Scholar]
- 4.Tucker JS, Ryan G, Golinelli D, et al. Substance Use and Other Risk Factors for Unprotected Sex: Results from an Event-Based Study of Homeless Youth. AIDS Behav. doi: 10.1007/s10461-011-0017-9. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Clatts MC, Goldsamt L, Yi H, et al. Homelessness and Drug Abuse among Young Men Who Have Sex with Men in New York City: A Preliminary Epidemiological Trajectory. J Adolesc. 2005;28:201–214. doi: 10.1016/j.adolescence.2005.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kipke MD, Weiss G, Wong CF. Residential Status as a Risk Factor for Drug Use and HIV Risk among Young Men Who Have Sex with Men. AIDS Behav. 2007;11:56–69. doi: 10.1007/s10461-006-9204-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cochran BN, Stewart AJ, Ginzler JA, et al. Challenges Faced by Homeless Sexual Minorities: Comparison of Gay, Lesbian, Bisexual, and Transgender Homeless Adolescents with Their Heterosexual Counterparts. Am J Public Health. 2002;92:773–777. doi: 10.2105/ajph.92.5.773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gangamma R, Slesnick N, Toviessi P, et al. Comparison of HIV Risks among Gay, Lesbian, Bisexual and Heterosexual Homeless Youth. J Youth Adolesc. 2008;37:456–464. doi: 10.1007/s10964-007-9171-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rew L, Whittaker TA, Taylor-Seehafer MA, et al. Sexual Health Risks and Protective Resources in Gay, Lesbian, Bisexual, and Heterosexual Homeless Youth. J Spec Pediatr Nurs. 2005;10:11–19. doi: 10.1111/j.1088-145x.2005.00003.x. [DOI] [PubMed] [Google Scholar]
- 10.Ennett ST, Bailey SL, Federman EB. Social Network Characteristics Associated with Risky Behaviors among Runaway and Homeless Youth. J Health Soc Behav. 1999;40:63–78. [PubMed] [Google Scholar]
- 11.Tyler KA. Social Network Characteristics and Risky Sexual and Drug Related Behaviors among Homeless Young Adults. Soc Sci Res. 2008;37:673–685. doi: 10.1016/j.ssresearch.2007.09.004. [DOI] [PubMed] [Google Scholar]
- 12.Wenzel SL, Holloway I, Munjas B, et al. Social Networks of Homeless Youth in Emerging Adulthood. J Youth and Adolesc. doi: 10.1007/s10964-011-9709-8. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bailey SL, Camlin CS, Ennett ST. Substance Use and Risky Sexual Behavior among Homeless and Runaway Youth. J Adolesc Health. 1998;23:378–388. doi: 10.1016/s1054-139x(98)00033-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kipke MD, O'Connor S, Palmer R, et al. Street Youth in Los Angeles Profile of a Group at High Risk for Human Immunodeficiency Virus Infection. Arch Pediatr Adolesc Med. 1995;149:513–519. doi: 10.1001/archpedi.1995.02170180043006. [DOI] [PubMed] [Google Scholar]
- 15.Rice E, Milburn NG, Rotheram-Borus MJ. Pro-Social and Problematic Social Network Influences on HIV/AIDS Risk Behaviors among Newly Homeless Youth in Los Angeles. AIDS Care. 2007;19:697–708. doi: 10.1080/09540120601087038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen X, Tyler KA, Whitbeck LB, et al. Early Sexual Abuse, Street Adversity, Ans Drug Use among Female Homeless and Runaway Adolescents in the Midwest. J Drug Issues. 2004;34:1–22. [Google Scholar]
- 17.Darling N, Palmer RF, Kipke MD. Do Street Youths' Perceptions of Their Caregivers Predict HIV-Risk Behavior? J Fam Psychol. 2005;19:456–464. doi: 10.1037/0893-3200.19.3.456. [DOI] [PubMed] [Google Scholar]
- 18.Ennett ST, Federman EB, Bailey SL, et al. HIV-Risk Behaviors Associated with Homelessness Characteristics in Youth. J Adolesc Health. 1999;25:344–353. doi: 10.1016/s1054-139x(99)00043-9. [DOI] [PubMed] [Google Scholar]
- 19.Rohde P, Noell J, Ochs L, et al. Depression, Suicidal Ideation and Std-Related Risk in Homeless Older Adolescents. J Adolesc. 2001;24:447–460. doi: 10.1006/jado.2001.0382. [DOI] [PubMed] [Google Scholar]
- 20.Liu H, Feng T, Feng H, et al. Egocentric Networks of Chinese Men Who Have Sex with Men: Network Components, Condom Use Norms, and Safer Sex. AIDS Patient Care ST. 2009;23:885–893. doi: 10.1089/apc.2009.0043. [DOI] [PubMed] [Google Scholar]
- 21.Tobin KE, Latkin CA. An Examination of Social Network Characteristics of Men Who Have Sex with Men Who Use Drugs. Sex Transm Infect. 2008;84:420–424. doi: 10.1136/sti.2008.031591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wenzel SL, Tucker JS, Golinelli D, et al. Personal Network Correlates of Alcohol, Cigarette, and Marijuana Use among Homeless Youth. Drug Alcohol Depend. 2010;112:140–149. doi: 10.1016/j.drugalcdep.2010.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Perreira KM, Deeb-Sossa N, Harris KM, et al. What Are We Measuring? An Evaluation of the CES-D across Race/Ethnicity and Immigrant Generation. Soc Forces. 2005;83:1567–1602. [Google Scholar]
- 24.Radloff LS. The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Appl Psychl Meas. 1977;1:385–401. [Google Scholar]
- 25.Bogart LM, Kral AH, Scott A, et al. Condom Attitudes and Behaviors among Injection Drug Users Participating in California Syringe Exchange Programs. AIDS Behav. 2005;9:423–432. doi: 10.1007/s10461-005-9014-1. [DOI] [PubMed] [Google Scholar]
- 26.Jemmott LS, Jemmott JB., III Applying the Theory of Reasoned Action to AIDS Risk Behavior: Condom Use among Black Women. Nurs Res. 1991;40:228–234. [PubMed] [Google Scholar]
- 27.Rice E. The Positive Role of Social Networks and Social Networking Technology in the Condom-Using Behaviors of Homeless Young People. Public Health Rep. 2010;125:588–595. doi: 10.1177/003335491012500414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Doll LS, Byers RH, Bolan G, et al. Homosexual Men Who Engage in High-Risk Sexual Behavior. A Multicenter Comparison. Sex Transm Dis. 1991;18:170–175. doi: 10.1097/00007435-199107000-00009. [DOI] [PubMed] [Google Scholar]
- 29.A Survey of AIDS Knowledge, Attitudes and Behaviours in San Francisco’s American-Indian, Filipino and Latino Gay and Bisexual Males Communities. San Francisco, California: San Francisco Department of Public Health, AIDS Office; 1991. Fairbank, Bregman, and Maullin [consulting firm] [Google Scholar]
- 30.Lemp GF, Hirozawa AM, Givertz D, et al. Seroprevalence of HIV and Risk Behaviors among Young Homosexual and Bisexual Men. The San Francisco/Berkeley Young Men's Survey. JAMA. 1994;272:449–454. [PubMed] [Google Scholar]
- 31.NTFAP. The 1991 Southern States AIDS Education Survey: A Knowledge, Attitudes and Behavior Study of African-American, Latino and White Men. Final Report: National Task Force on AIDS Prevention. 1993. [Google Scholar]
- 32.Richwald GA, Morisky EE, Kyle GR, et al. Sexual Activities in Bathhouses in Los Angeles County: Implications for AIDS Prevention Education. J Sex Res. 1988;24:169–180. [Google Scholar]
- 33.Mimiaga MJ, Noonan E, Donnell D, et al. Childhood Sexual Abuse Is Highly Associated with HIV Risk-Taking Behavior and Infection among Msm in the Explore Study. J Acquir Immune Defic Syndr. 2009;51:340–348. doi: 10.1097/QAI.0b013e3181a24b38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Paul JP, Catania J, Pollack L, et al. Understanding Childhood Sexual Abuse as a Predictor of Sexual Risk-Taking among Men Who Have Sex with Men: The Urban Men's Health Study. Child Abuse Negl. 2001;25:557–584. doi: 10.1016/s0145-2134(01)00226-5. [DOI] [PubMed] [Google Scholar]
- 35.Tucker JS, Edelen MO, Ellickson PL, et al. Running Away from Home: A Longitudinal Study of Adolescent Risk Factors and Young Adult Outcomes. J Youth Adolesc. 2011;40:507–518. doi: 10.1007/s10964-010-9571-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Yates GL, Mackenzie RG, Pennbridge J, et al. A Risk Profile Comparison of Homeless Youth Involved in Prostitution and Homeless Youth Not Involved. J Adolesc Health. 1991;12:545–548. doi: 10.1016/0197-0070(91)90085-z. [DOI] [PubMed] [Google Scholar]
- 37.Aidala A, Cross JE, Stall R, et al. Housing Status and HIV Risk Behaviors: Implications for Prevention and Policy. AIDS Behav. 2005;9:251–265. doi: 10.1007/s10461-005-9000-7. [DOI] [PubMed] [Google Scholar]
- 38.Wolitski RJ, Kidder DP, Fenton KA. HIV Homelessness, and Public Health: Critical Issues and a Call for Increased Action. AIDS Behav. 2007;11:167–171. doi: 10.1007/s10461-007-9277-9. [DOI] [PubMed] [Google Scholar]
- 39.Amirkhanian YA, Kelly JA, Kabakchieva E, et al. Evaluation of a Social Network HIV Prevention Intervention Program for Young Men Who Have Sex with Men in Russia and Bulgaria. AIDS Educ Prev. 2003;15:205–220. doi: 10.1521/aeap.15.4.205.23832. [DOI] [PubMed] [Google Scholar]
- 40.Young SD, Rice E. Online Social Networking Technologies, HIV Knowledge, and Sexual Risk and Testing Behaviors among Homeless Youth. AIDS Behav. 2011;15:253–260. doi: 10.1007/s10461-010-9810-0. [DOI] [PMC free article] [PubMed] [Google Scholar]