|Home | About | Journals | Submit | Contact Us | Français|
African Americans face disproportionately higher risks of HIV infection. Concurrent sexual relationships, or sexual partnerships that overlap in time, are more common among African Americans than individuals of other races and may contribute to racial disparities in HIV infection. However, little is known about attitudes, norms and practices among individuals engaged in concurrent partnerships. Little is also known about the processes through which structural, behavioral and social factors influence concurrent sexual relationships. We recruited 24 heterosexual African American men involved in concurrent sexual relationships from a public health clinic in Philadelphia. We conducted in-depth interviews exploring these men's sexual practices; social norms and individual attitudes about concurrency; perceived sexual health risks with main and non-main partners; and the social, structural and behavioral factors contributing to concurrent sexual relationships. Twenty-two men reported having one main and one or more non-main partners; two reported having no main partners. Respondents generally perceived sexual relationships with non-main partners as riskier than relationships with main partners and used condoms far less frequently with main than non-main partners. Most participants commented that it is acceptable and often expected for men and women to engage in concurrent sexual relationships. Social factors influencing participants’ concurrent partnerships included being unmarried and trusting neither main nor non-main partners. Structural factors influencing concurrent partnerships included economic dependence on one or more women, incarceration, unstable housing, and unemployment. Several men commented that individual behavioral factors such as alcohol and cocaine use contributed to their concurrent sexual partnerships. Future research and interventions related to sexual concurrency should address social and structural factors in addition to conventional HIV risk-taking behaviors.
African Americans’ HIV infection rates are seven times those of Whites (Hall et al., 2008). While African Americans represent 13% of the US population, they account for 45% of new HIV infections (Hall, et al., 2008). Men account for two-thirds (65%) of African Americans’ new infections (Hall, et al., 2008). Nearly 40% of HIV infections among African American men are among heterosexuals (Hall, et al., 2008); this likely contributes to high infection rates among African American women (CDC, 2004).
Traditional behavioral risk factors including condom use, number of sexual partners, and substance use do not fully explain racial disparities in HIV infection (Hallfors, Iritani, Miller, & Bauer, 2007). Social determinants of HIV risk, defined broadly as “the conditions in which people are born, grow, live, work and age, including the health system” (WHO, 2008) may be equally as important as individual behavioral factors in explaining these racial disparities (Aral, Adimora, & Fenton, 2008).
New literature suggests that African Americans’ sexual networks are strongly socially determined and contribute to HIV risks (Aral, et al., 2008; Laumann & Youm, 1999). Concurrent sexual partnerships, or partnerships that overlap in time, raise HIV transmission risks more than simply having numerous consecutive, monogamous sexual partnerships (Adimora, Schoenbach, & Doherty, 2006; Garnett & Johnson, 1997; Martina Morris & Kretzschmar, 1995; M. Morris & Kretzschmar, 1997) (Kretzschmar, 2000). Concurrency may contribute to HIV infection among African Americans (Adimora et al., 2002; Adimora, Schoenbach, & Doherty, 2007; Adimora et al., 2004) and has been associated with HIV infection in several studies (Adimora, et al., 2004; Adimora et al., 2006; Adimora et al., 2003). Recent analyses found African American men had 2.56 higher odds of engaging in concurrency as White men (Adimora, et al., 2007).
For the purposes of our analyses, we unpack social and individual determinants of health into distinct categories. We define social factors as “factors related to how humans interact and relate to others,” structural factors as “physical, environmental or economic factors” (WHO, 2008) and behavioral factors as “individual behaviors that influence HIV risk.” Concurrency has been statistically associated with social factors such as single marital status (Adimora, et al., 2004), and individuals engaged in concurrency often report believing their partners are also engaged in concurrency (Adimora, et al., 2003; Gorbach, Stoner, Aral, WL, & Holmes, 2002; Magnus et al., 2009). Concurrency has been associated with structural factors such as incarceration (Adimora, et al., 2004; Adimora, et al., 2003; Khan et al., 2009a; Manhart, Aral, Holmes, & Foxman, 2002), poverty (Adimora, et al., 2002; Adimora, et al., 2007), and behavioral factors including drug and alcohol use (Adimora, et al., 2007; Adimora, Schoenbach, Taylor, Khan, & Schwartz, 2011). However, little is known about practices, attitudes and social norms among individuals engaged in concurrent partnerships. The processes through which individual and social determinants of health influence concurrent partnerships are also not yet well understood. This qualitative analysis helps fill that gap.
Philadelphia's HIV incidence rates are five times the national average (AACO, 2008). Nearly 70% of new HIV infections in Philadelphia in 2006 were among African Americans (AACO, 2008); two percent of African Americans in Philadelphia are HIV-positive (Schwartz, Feyler, Baker, & Brady, 2010). Philadelphia recently implemented a rapid HIV testing program in public health clinics; since 2007, approximately 15,000 people have undergone rapid HIV testing; 88% are African American (Nunn et al., 2010).
We conducted in-depth interviews with African American men engaged in concurrent sexual partnerships to explore attitudes and practices about concurrency and to unpack the social, structural and behavioral factors influencing this important risk factor for HIV transmission.
We recruited 24 heterosexual African American men who reported concurrent partnerships when testing for HIV in a Public Health Center in a high-incidence zipcode in Philadelphia. Recruitment took place twice monthly during 2009. All patients participated in behavioral risk assessments before testing. Eligibility criteria included: engaging in one or more concurrent partnerships, being 18 or older, having undergone rapid HIV testing within the last month, self-identifying as African American and heterosexual, reporting only ever having sex with women, and providing written informed consent.
We employed the United Nations Working Group on Concurrent Partnerships definition of concurrency: “overlapping sexual partnerships in which sexual intercourse with one partner occurs between two acts of intercourse with another partner within the last six months” (UNAIDS, 2010).
Medical assistants offered patients an opportunity to participate after conducting their rapid HIV tests. Declining to participate did not affect participants’ clinical care. Individuals provided informed consent. Study participants received a $30 Wal-mart™ card and a $20 transportation incentive. Protocols were approved by Lifespan and Philadelphia Department of Public Health Institutional Review Boards.
We used the “grounded theory” qualitative interviewing approach in which data collection informs development of theory building and subsequent data analysis (Glaser & Strauss, 1967). A semi-structured interview guide was informed by literature on social, (Adimora, et al., 2004; Adimora, et al., 2003; Magnus, et al., 2009) structural (Adimora, et al., 2004; Adimora, et al., 2003; Khan, et al., 2009a; Manhart, et al., 2002)(Adimora, et al., 2002; Adimora, et al., 2007), and behavioral factors influencing concurrent partnerships (Adimora, et al., 2007; Adimora, et al., 2011) and data on Philadelphia's HIV/AIDS epidemic.(Schwartz, Feyler, Baker, & Brady, 2009; Schwartz, et al., 2010) Guides included questions about participants’ sexual practices and attitudes about concurrency, social norms about concurrency, why participants engage in concurrency, and condom use practices with main and non-main partners. We defined main partners as partners with whom the respondents had an emotional bond and with whom they had regular sexual intercourse, such as girlfriend, spouse, significant other, or life partner. We defined non-main partners as partners with whom respondents had occasional or one-time only sexual intercourse. Guides inquired about how social factors such as marriage, availability of partners and support systems; structural factors such as incarceration, employment and poverty; and individual behavioral factors such as drug and alcohol use impacted concurrency. Interviews lasted 45 to 75 minutes and were loosely structured to allow interviewer and respondents to freely introduce topics (Seidman, 1998; Weiss, 1994).
Interviews were recorded and professionally transcribed. Transcripts were coded to identify themes. Open coding permitted grouping of themes according to common topics that arose during interviews, including those not in interview guides (Rice & Ezzy, 2001; Seidman, 1998; Weiss, 1994). To ensure reliability and validity of findings, approximately 20% of interviews were coded by multiple data analysts; codes were checked for concordance. Analytic memos were drafted to summarize and link key interview themes; memos informed study findings.
Table 1 includes demographic information about participants. Many were unemployed (14/24) and lacked stable housing (11/24); two were married. Most had been incarcerated (17/24) and reported main and non-main partnerships (22 /24); two had no main partners. Most reported differential condom use among main and non-main partners and that trusting their partners affected concurrency (Table 2).
Four primary themes emerged: 1) attitudes and practices related to concurrency; 2) social 3) structural and 4) individual behavioral factors influencing concurrent relationships. Employing quotes exemplifying emergent themes, we first explain respondents’ attitudes and practices related to concurrency and then present key factors impacting their concurrent partnerships.
Most men had main partners with whom they shared an emotional bond or long-term relationship and non-main partners that played different roles in their lives. Two respondents explained:
The one I was really close to worked at Social Services. The other three were weekend and whenever [partners]. They'd pop over to have sex... each did something the others didn't. One of them had a bunch of money. The other one was a freak. Anything [sexual] goes for her. The other one had my back with cops.
I had to make this chart. To separate them. Pros and cons for each. There are a lot of different things about each of them. A couple of them I like because they take care of me. Two or three of them spoil me... when I come over, its all about me! A couple of them like to argue a lot, but the sex is good. Then you got some that cook good, then a couple that talk to me a certain way that I like. They motivate me.
In contrast, one man described why he had no main partner:
I don't want a main girlfriend or any questions or problems, anybody looking through my cell phones, just stopping by my house. I don't want any of that.
Many respondents reported using condoms less frequently, if at all, with main partners. Respondents generally perceived sexual relationships with non-main partners as riskier than with main partners. Several alluded to risks associated with multiple partners, but continued to engage in concurrency. One individual who never used condoms with his main partner explained condom use with non-main partners:
If I don't feel right about her, if I think she needs a condom, I ask. I peruse females’ characters. If I'm in a conversation and they mention sex before I do, when they bring it up, it makes me think it's not the only conversation they have had with males. Because I don't mention sex at all! So if you're coming off kind of strong, I'm going to look at what you do.
Many men commented it was socially acceptable, and even expected, for men to have multiple concurrent partners. Many associated concurrency with masculinity.
I don't know what it is – it's just a thing. A male thing I guess. You feel as though you've got to be dominant.
It's not cute for a female to run around having a bunch of partners. That's not ladylike to me. It's different for guys. Females like that on the streets are considered sluts, whores... But you ain't ever heard anybody say a man is a whore.
Men often commented their partners knew about their concurrent partnerships:
They [non-main partners] talk to me, tell me their problems and stuff. But they know I've got a girlfriend. They don't like it because they want to be my girlfriend, but they've got to deal with it... If I don't lie to them, then it's easy. I'm open about it.
Numerous respondents believed their female partners often had concurrent partners:
I just know there is more than just me because I'm not there all the time. You know there's somebody else. You don't constantly take your car to one garage-- you got to take it to another.
On the surface you don't care because it's [concurrency] an everyday thing. You are used to it and immune to really being scared of getting something.
Numerous respondents engaged in concurrency because they didn't trust their partners not to engage in concurrent relationships. One respondent explained why he had concurrent partners:
Only because they [my partners] were active. If I knew that person was cheating, I'd cheat on that person. I don't call it cheating because we told each other.
Many men's lack of trust of partners was linked to general distrust of people in the community. One man's comment reflected this common sentiment:
Can't really trust them [sexual partners], like all people in the streets, they're not to be trusted. You can only trust them as far as you can see them...My family-- I don't have to worry about what they're doing--- But anybody else, I don't trust...In general, people are like that.
Most participants commented that being unmarried contributed to concurrency. One man's remark reflects this common theme:
When I feel as though I'm ready to settle down, put that ring on her finger, that's when I'm going to stop all my bull and just be with her, but I'm not ready for that right now.
Most participants had been incarcerated (17/24) and explained how incarceration interrupted partnerships and affected decisions to initiate new relationships. Many also believed their female partners had other partners during their incarceration:
[I was incarcerated] once. That was my first [partner], my son's mom. After her, I wasn't really in any relationships. That's why it ended. Because she was having those relationships while I was in prison. I got locked up, then we faded and went our separate ways, and she was bringing the kids up for a while.
That's why it [the relationship] ended. Because she was having those relationships while I was in there. I did too much damage. When I went to do my last little bit [of my prison sentence].... Things were really bad and we went our separate ways.
Unemployment prompted many men to depend economically on concurrent partners:
I've got this person, that person. If I need transportation money you go over here; if I need this you go over here; if I need a place to stay, I go over there.
Many unemployed men depended on partners for housing. One unemployed respondent explained:
Me and my mother are not getting along. So I was staying with my girlfriend but that didn't work out too well because she lives with her parents. I'm staying there now because I don't have anywhere else to go, but I'm trying to find somewhere.
Numerous men commented they sometimes gave and received gifts or favors from concurrent partners:
I have received things from women-- material things-- and I haven't had to return material things to them.
Drug use was frequently linked to concurrency, particularly one-time sexual encounters. Nine respondents reported cocaine influenced concurrency. One respondent's comment reflected this common theme:
My drug of choice was crack cocaine. It raises your sexual level. You get a higher sex drive. The more you crave coke, the more you go out and get sex. I don't really think too much about doing the right thing. I just do whatever works for me at the time. Whatever means necessary. I don't really think “We need to go put a condom.”
Several respondents reported regular alcohol and marijuana use. When asked whether alcohol and marijuana impacted concurrency, responses varied widely. Four individuals reported alcohol use impacted concurrency, while one man who reported consuming four to five alcoholic beverages daily stated alcohol did not affect sexual decisions. Another reported his marijuana use did not impact sexual decision-making but noted alcohol's strong effect:
Alcohol is something else –for me, it just makes me not care. I could go home with anybody when I'm drunk, with any girl.
Social, structural and behavioral factors influenced concurrent sexual partnerships among these low-income African American men, suggesting that social determinants of HIV risk may be equally important as individual behavioral factors in influencing concurrent partnerships. Most participants had main and non-main partners and indicated they relied on their main partner for emotional support and companionship while relying on other partners for housing, economic support, or sex. This finding is supported by another study that found concurrent partners played different roles in men's lives (Carey, Senn, Seward, & Vanable, 2008).
Structural factors such as poverty, unemployment and economic dependence on women for housing and food strongly impacted concurrent relationships. Philadelphia has the highest poverty rate among the nation's ten largest cities (Butkovitz, 2010; “Philadelphia: The State of the City,” 2010), and most participants are from poor inner-city communities that had unemployment levels above 10% prior to the 2008 economic downturn (BLS, 2004). Other literature associates living below the poverty line with concurrent partnerships among men and women (Adimora, et al., 2002; Adimora, et al., 2007). However, little research explores how economic dependence on partners contributes to concurrency; our findings linking concurrency to unemployment and unstable housing represent novel contributions to the concurrency literature.
Most participants (17/24) had a history of incarceration; numerous men noted incarceration interrupted their partnerships and influenced their partners’ concurrency. Philadelphia has the nation's 4th highest incarceration rate; 5.7 of every 1,000 residents are behind bars (Eichel, 2010). Our findings noting incarceration's impact on concurrency add to a growing body of research documenting associations between incarceration of African American men, concurrency and interruption of sexual partnerships (Adimora, et al., 2004; Adimora et al., 2001; Aral, et al., 2008; Khan et al., 2009b; Khan, Miller, et al., 2008; Khan, Wohl, et al., 2008).
Marital status and social norms were important social factors influencing concurrent sexual partnerships. Nearly 45% of African Americans in Philadelphia have never been married; (BLS, 2004) fewer than 10% of individuals living below the poverty line in Philadelphia are married (BLS, 2004). Engaging in concurrent sexual relationships when one is unmarried was socially acceptable, expected, and associated with masculine identity among this population. Notably, many respondents commented they expected their female partners to engage in concurrent relationships. Taken together, our findings suggest concurrency has been normalized in this community with low marriage rates; this is a novel contribution of this study.
The most common reason participants cited for engaging in concurrency was trusting neither main nor non-main partners. This phenomenon has not been explored elsewhere in the concurrency literature. Notably, many respondents believed their partners had other partners, and noted this affected their own decisions to pursue concurrent relationships, often in spite of perceived health risks. Participants generally reported trusting main partners more than non-main partners. Participants also noted that trust impacted condom use practices; most men in our sample rarely, if ever, used condoms with their main partners and only sometimes used condoms with non-main partners. Other studies also report more consistent condom use with non-main than main partners (Fortenberry, Tu, Harezlak, Katz, & Orr, 2002; Howard, Fortenberry, Blythe, Zimet, & Orr, 1999; Misovich, Fisher, & Fisher, 1997; Rosengard et al., 2001). Inconsistent condom use among concurrent partners likely increases HIV transmission.
Many also reported that distrusting their community affected marital decisions and concurrent partnerships. Low levels of trust of sexual partners and of the community at large may reflect low levels of trust and social capital in this community. Social capital is defined as levels of “community trust, community participation and civicness” (Putnam, 1993, 2000). High crime rates, low marriage rates, and high poverty rates are frequently associated with low levels of neighborhood social capital, low socioeconomic status (SES) and poor health outcomes (Berkman & Kawachi, 2000; Kawachi, Kim, Coutts, & Subramanian, 2004; Saegert, Thompson, & Warren, 2001; Subramanian & Kawachi, 2006). Not trusting others, including sexual partners, suggests low levels of social capital may have important impacts on concurrent sexual partnerships; these important phenomena are fertile ground for ongoing research about social determinants of concurrent sexual relationships.
Additionally, substance use, and cocaine use in particular, was cited as a contributing factor to concurrency. This supports other research associating drug and alcohol use with increased HIV risk behaviors (Donovan & McEwan, 1995; Leigh, 2002; Leigh et al., 2008; Sheth, Moore, & Gebo, 2006; Woolf & Maisto, 2009).
Although addressing concurrency's social determinants presents formidable challenges, one prime opportunity for interventions to address concurrent partnerships may be reducing drug-related incarceration. African American men are incarcerated for drug offenses at disproportionately higher rates (Western, 2005), and racial disparities in incarceration have been correlated with HIV/AIDS prevalence among African Americans (Johnson & Raphael, 2009). Policies to reduce prison sentences for drug-related crimes, drug-related recidivism, parole violations and harm reduction policies could be less socially disruptive than incarceration, and may ultimately reduce concurrency and communal HIV risks.
Our study is subject to several limitations. We interviewed small sample of African American men in Philadelphia who reported engaging in concurrent sexual partnerships, and did not recruit men of other races from similar socioeconomic strata. Findings may not represent broader trends among heterosexual men or even heterosexual African American men; rather, these findings reflect the experiences of a group of African American men of low SES engaged in concurrency.
However, this study underscores the importance of social determinants of HIV risks and concurrent partnerships among low-income African American men. Individual-level behavioral interventions may overlook the important role of social networks in potentiating health disparities in HIV/AIDS and may fail to address the critical roles of social and structural factors that contribute to concurrent sexual partnerships.
This manuscript was supported by grant numbers K01 AA020228-01A1 from the National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health (NIAAA/NIH); grant number P30-AI-42853 from the National Institutes of Health, Center for AIDS Research (NIH/CFAR); training grant number 5T32DA13911 from the National Institute on Drug Abuse, National Institutes of Health (NIDA/NIH); and a Rhode Island Foundation Medical Research Grant. None of the aforementioned agencies had any role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
We thank Drew de Los Reyes and Gladys Thomas for their comments on previous drafts of this manuscript.
Amy Nunn receives consulting fees from Mylan Inc.