Rates of HIV infection are rapidly increasing among women in the United States and worldwide.^1,2 The majority of new infections are among women who contracted the virus from their primary male sexual partners;³ the predominance of this route of transmission reflects that condom use is lower in primary relationships than in casual or paying relationships.^4–6 A variety of factors are associated with couples’ use of condoms, including gender inequity and differential power relations;^7–9 occurrence of intimate partner violence;^10,11 levels of perceived risk^12,13 and self-efficacy;^14,15 and beliefs that condoms feel unnatural, reduce sexual pleasure and are a hindrance to intimacy.^16,17

Some studies have focused on women's reluctance to use condoms in primary relationships and have included reproductive intentions^9,18 or the use of hormonal and other forms of contraception.¹² Other research has focused on the emotional and social needs that are fulfilled in committed relationships, and has found that condom use may be inconsistent with relationship ideals of intimacy, trust and fidelity.^9,19 These explanations may be particularly relevant for inconsistent condom use by couples who know that they are serodiscordant, but not for those in which partners have the same HIV status.^3,20 Further, most research on the relationship between emotional needs, romantic attachment and condom use has studied the importance of primary relationships from the woman's perspective but has overlooked the man's.

Public health policymakers and practitioners conceptualize sex in terms of disease and unwanted pregnancy, and not in terms of love, intimacy and trust. This is particularly the case with regard to substance users, who typically are described as having sex partners, rather than spouses or lovers.²¹ Regardless of the population, HIV prevention efforts have focused primarily on the individual, emphasizing risk reduction through safer sex, often neglecting the context in which risky behavior occurs. However, safer sex in the age of HIV and AIDS entails talking about condoms and acknowledging the possibility of disease. Such discussion may not be conducive to a romantic, spontaneous, passion-filled encounter that might lead to love and a long-term relationship. In contrast, some people use casual or unprotected sex as a strategy to "catch love."²² Among heterosexuals, particularly young women, trust and love are central in defining sexual involvements. Young women, and sometimes men, often "trust to love," in that they see condom use as unnecessary with a regular partner or within a relationship.^16,23 Furthermore, young women often view love and trust as prophylactic, and believe that sex is safe because of its relationship with love.²²

In the search for love and a meaningful relationship, people may not always act rationally. For example, they may base decisions about sex and condom use on implicit personality- or characteristics-based risk theories (e.g., on whether a potential partner is from the streets or looks clean).^24,25 Also, once a relationship is established, men and women may follow one of several inaccurate AIDS prevention rules of thumb that support nonuse of condoms (e.g., "Known partners are safe partners" or "It's too late to start using protection now").²⁴ Condoms and their use within a primary relationship, particularly an established relationship in which they may not have been used previously, can raise issues of distrust and accusations of infidelity.^8,26 Unprotected sex helps maintain the fantasy that one's partner is faithful.²⁷

In recent years, negotiated safety as a risk assessment and reduction strategy has been a focus of research.^28–30 Negotiated safety refers to an explicit agreement between partners about sexual practices that takes into account the HIV status of both. Couples using this approach undergo mutual HIV testing, then decide to discontinue condom use within their primary relationship and commit to monogamy or establish rules for condom use with outside partners. While the frequent use of negotiated safety among homosexual men is well documented,³¹ its use by heterosexual couples has received little attention, and researchers disagree as to whether heterosexuals are likely to employ it effectively.^22,32

In this article, we examine condom use within the context of high-risk urban couples, including both HIV-seroconcordant and HIV-serodiscordant couples. Specifically, we explore reasons cited for condom use or nonuse, and the strategies used to reduce STD risk, within primary partnerships.

METHODS

Our data were collected as part of a study conducted in Hartford in 2004–2007, exploring factors affecting high-risk men and women's use of the female condom for STD prevention after barriers to awareness and accessibility have been removed. High-risk individuals were defined as illicit drug users, partners of injection-drug users, commercial sex workers, and homeless or poor individuals. In addition to longitudinal surveys conducted with a cohort of more than 400 women, the study included a couples component of qualitative and quantitative longitudinal interviews and a two-week trial of the female condom. We present findings from the baseline in-depth narrative interviews conducted with 25 couples.

Recruitment for the couples component of the study used two approaches. Initially, every third woman participating in the larger cohort study was screened for eligibility and was asked if she would like to participate and would be able to bring in her male sexual partner. Because this approach proved difficult, we changed to direct recruitment of couples using targeted outreach. Study staff approached men and women found together in parks and on the street outside of social service organizations, and asked them if they were a couple. If they responded positively, staff told them about the study, screened them for eligibility and scheduled them for an appointment to enroll at the study office. To be eligible, individuals had to be 18 or older, live in the Hartford area and have had vaginal sex with this partner during the last 30 days.

Participants in the couples study provided a urine sample for screening for chlamydia and gonorrhea. If either partner reported current STD symptoms during the survey or tested positive with the urine screen, the couple was excluded from participation in this component of the study and was referred to the local health department for treatment.^* In-depth narrative interviews, conducted at baseline and 10 months, explored current and past relationships; relationship ideals; sexual behavior; sexual and contraceptive decision making; STD risk and preventive practices; violence; and knowledge of, attitudes toward and experiences with the female condom. Men and women were always interviewed separately and were compensated $25 for each in-depth interview. Written informed consent was obtained from each participant prior to STD screening and study enrollment. All study protocols were reviewed and approved by The Institute for Community Research institutional review board prior to initiation of any research activities.

All qualitative interviews were transcribed verbatim, and text data were coded and analyzed through an iterative inductive and deductive process using Atlas.ti software. Interviews were first coded for demographic variables. Data were then coded for content using a coding scheme developed for the project, which comprised the following broad categories: first sexual experience with current partner, male condom use, general attitudes toward condoms, reproductive intentions, sexual negotiation, sexual practices, sexuality and pleasure, violence and substance use. In a third level of analysis, members of the research team worked jointly to identify key themes and patterns of response and relationships among the variables of interest. For example, the importance of love was a recurrent theme and was related to many other codes, such as first sexual experience with partner, sexual negotiation and sexual practices.

RESULTS

The majority of participants were nonwhite and poor, and had low levels of educational attainment. Nearly half currently used crack or injection drugs, and the majority had abused these drugs in the past. Eight couples were homeless at the time of their first interview. Six women and seven men reported that they were HIV-positive; three couples were seroconcordant positive, and the other seven were serodiscordant. Current relationship duration averaged five years (standard deviation, 59.7 months; range, 1–240 months). The overwhelming majority of couples, including six of the 10 affected by HIV, reported never using condoms.

Two main themes emerged from the in-depth interviews: Couples often feel that the love and intimacy needs that romantic relationships meet supersede health concerns, and many couples reduce their STD risk through negotiated safety and similar strategies.

Love and Intimacy

As they described the beginning of their current relationships and first sexual encounters, both women and men talked about "taking a chance on love." Nonuse of condoms was a strategy employed to help establish a potentially serious, as opposed to casual, relationship. Amanda (a white, HIV-negative 41-year-old)^† had known Juan for two weeks before they first had sex. They met while in an inpatient detoxification program, and she described both an immediate physical attraction and a deeper emotional bond stemming from hours of conversation about their experiences. At the time of their baseline interviews, they had been together for about three months. Explaining why they chose not to use condoms during their first sexual experience, Amanda stated:

"I knew I was just going to be with him. He wasn’t just going to be this one person I was just going to mess around with and then … not see him. I was at least hoping that I was going to be with him [for a long time]."

Amanda's desire to be in a long-term relationship and belief that Juan could be the one made it easier not to use condoms during their first sexual encounter. In turn, their nonuse of condoms supported the hope that this was not just a fling, but rather a first experience with someone who was equally invested in fostering a meaningful primary relationship.

Several participants, in describing their ideal romantic partner, talked about having spent years searching for a soul mate and having found one in their current partner. Carmen (a Hispanic, HIV-negative 44-year-old) had spent several years in an abusive relationship before she met Manuel, who was involved in a relationship at the time. While they were physically attracted to each other, they spent a year as friends; during that time, their feelings deepened. Carmen's realization that Manuel might be her soul mate was somewhat complicated by the fact that he was HIV-positive. However, the importance of finding a life partner, particularly a nonabusive one, outweighed her fears about health risks as she started a sexual relationship with him. Carmen and Manuel had been together about eight years and used condoms inconsistently. Carmen explained:

"I'll be honest; I was a little bit scared because I didn’t know much about AIDS or anything like that. … [But] the chemistry. … We’re supposed to be together, a soul mate. It's like something that you have together; that's how it was with me and him. … Being sick or whatever can’t stop that from happening. You understand what I mean?"

For many participants, the first sexual encounter with their current partner was not just about sex but also about establishing trust. In these situations, having sex without a condom was used as evidence that a partner had been honest about his or her sexual past and disease status, or was not currently involved in another sexual relationship. Veronica (a Hispanic, HIV-negative 38-year-old) met Wilfredo while in methadone maintenance treatment. They became friends and started dating after two years. They had been together about two months at the time they enrolled in the study. Veronica recalled that as their relationship evolved, she "didn’t want to have sex right away," and she was as nervous as a teenager when they finally did. However, she continued:

"The first time … was so great, I could not believe it. No, we didn’t use [a condom]. And we didn’t even talk about it, but the day after, I ask him, ‘How come you didn’t use any protection with me?’ You know, because he had a girl before me, and he was using protection [with her]. He says, ‘Because I trust you, and I know you’ve been honest to me so far and I can trust you.’… He's the type of guy that when he's in love with a woman, he don’t like to use protection; but if he's just dating someone like he was dating this girl just for sex, he was using protection. So … I was so proud of him that at least he was using rubbers with them and the reason he didn’t with me [was] because he trusted me."

Veronica echoed many other participants’ view that the use of condoms signifies a level of distrust. In addition, many participants believed that condoms are appropriate for casual and "sex-only" encounters but not necessarily for primary, serious relationships.

Again, the need for physical and emotional intimacy outweighed health concerns. Veronica knew that Wilfredo had been having sex with another woman as they started to get to know each other, and she had no way of confirming whether he had used condoms with his former partner. However, she believed his assertion and took his nonuse of condoms with her as evidence of the seriousness of their relationship. Many women reported having unprotected sex despite some doubt about their partner's monogamy, choosing to believe that he was faithful. Jennifer (a white, HIV-negative 38-year-old) further illustrated this point in her remarks:

"We honestly believed that we wouldn’t ever have sex with anybody else again, which is still true this day. … That is one thing I do believe. And I do know that he's not messing around. I mean, well, in my heart I believe. I guess nobody can honestly say 100% that they believe their mate is messing around or not. … [But] we both had had HIV tests. He said his was negative. Mine was negative. So we just felt that in using [a condom], it would be more of a blocker of our love."

Doubt about a partner's fidelity did not typically lead to confrontation or insistence on condom use. Rather, it often resulted in continued unprotected sex, which supported the belief that the relationship was monogamous, safe and founded on true love. This, in turn, reinforced misperceptions about AIDS or people with AIDS and helped to maintain the fantasy of the ideal—namely, that a partner was faithful.²⁷

In addition to reduced physical pleasure, several participants associated using condoms with reduced intimacy, both physical and emotional. Martha (a black, HIV-negative 44-year-old) commented:

"What's the use of using [condoms]? I mean, Why? I think your body fluids or whatever is intimacy. … And the feeling, you know? His meat against my meat. … I think a condom means that you really don’t want contact. … It's only for hookers or somebody that you don’t … want to feel their immediate skin against yours."

Balancing the desire and need for physical intimacy and protecting oneself or one's partner was particularly difficult for serodiscordant couples, as illustrated by Manuel and Carmen. Manuel (a Hispanic, HIV-positive 33-year-old) explained that he had been using condoms for so many years, but sometimes he just "[wants] to feel, you know, flesh and flesh." He said that Carmen sometimes feels the same way, "so sometimes we have sex without a condom. … It's like something that we want to share, you know?"

Manuel's narrative also raises the issue of "condom fatigue," which is frequently discussed in research on sexual behavior among men who have sex with men³³ and may also operate in heterosexual—particularly serodiscordant—partnerships. Condom fatigue, or frustration with the prospect of having the pleasure of sex reduced at every encounter,³⁴ combined with the desire and need for physical intimacy, may prove overwhelming at times; as Manuel's comments suggest, it may even result in inconsistent condom use despite the very real risk of HIV transmission.

Stories related by participants in HIV-affected relationships illustrate that for many people, the emotional and other needs that are fulfilled by intimate relationships may supersede concerns for personal health and safety, leaving them unwilling to risk losing what they have. In extreme cases, they may be willing to die for love. Ana (a Hispanic, HIV-positive 37-year-old) had been with Lazaro for 20 years when she entered the study and had received an HIV diagnosis in 1992. Sometime before learning that she had HIV, she had found out from Lazaro's sister that he was HIV-positive, and had felt betrayed and hurt because he had not told her himself. Yet, she explained:

"[Knowing that Lazaro was infected] wouldn’t have changed nothing, just that … I would have been more protective. That's it, but I wouldn’t have left him because he had AIDS. Because the love was already there, you know? So I think maybe he was afraid. [Once I found out,] … I didn’t want the sex or nothing for a while, because I was hurt. … Then we started back again, and we didn’t use condoms. Never used condoms because I said, ‘Damn, I love him and everything. If I love him, I shouldn’t even use condoms with him.’ But I didn’t know [about reinfection] at the time, you know? Like I said, ‘That's my man, I love him and so what. We’ll die together.’ That's how I felt."

Ana's story also suggests additional themes. Feelings of betrayal at not only Lazaro's dishonesty regarding his HIV status but his disregard for her health are evident. Ana also displays a certain level of fatalism, or a belief that her destiny is not in her own hands.³⁵ Finally, once Ana had come to terms with the fact that the man she loved was HIV-positive, she had to decide whether to leave him, stay with him and use condoms, or stay with him and not use condoms. Although she chose to stay with Lazaro and not use condoms, it is not clear whether she did so because she felt that getting infected and dying with him would be better than being left alone after he died; because she assumed that he had already infected her, and she accepted the perspective that it was "too late to start using condoms now";²⁴ as a conscious expression of her love for him (which has been suggested as a major reason for nonuse of condoms in serodiscordant relationships²⁰); or for a combination of reasons.

Javier and Karla's story is another example of how love sometimes supersedes concern about risk. The couple had been together for three years at their first interview. She had received an HIV diagnosis in 2005, and Javier continued to test HIV-negative. When asked what prevents him from taking precautions against infection, Javier (a Hispanic 29-year-old) responded:

"That's a good question. I don’t know. I go with the flow like I always do. … Well, because I feel so comfortable and I feel great and I’m living this life, you know, and whatever happens, it happens, you know? Right now I’m enjoying it. Whatever time God wants [to give me and then] he’ll take me. I’m having the best time right now. I don’t want to change this for nothing."

When the interviewer probed this response, Javier confirmed that he is willing to risk his life "for being happy."

Assessing and Minimizing Risk

Several couples described assessing their HIV risk as they decided whether to use condoms, although the thoroughness with which they did so varied considerably, and some couples reported never explicitly discussing their risk or condom use. José and Linda, who had been together for four and a half years, decided to use protection at the beginning of their relationship because of their drug use histories. After assessing their risks and intentions for the relationship, they had HIV tests, agreed to be monogamous and engaged in unprotected sex. José, a Hispanic 32-year-old who was HIV-negative, related:

"When we first met, it took us a while before we had sexual relations, and we used condoms at first ’cause, you know, we met in a program. … Half of [the people there] are sick or something; they have some type of disease. So she had taken her [HIV test]. Well, she told me to use [condoms]. … She wanted to protect herself, you know? … She made the decision, and I agreed with it ’cause we just met not too long ago."

Asked at what point he and Linda had decided not to use any protection, José explained that after a while, they had agreed that since they both had been tested for HIV and other STDs and they were monogamous, they had no reason to use condoms. "We shouldn’t be sick," he said, "unless one of us went out there and fooled around and got sick."

In comparing José and Linda with other couples, we can see that negotiated safety may be viewed as a continuum of levels of risk assessment and potential responses to risk. Furthermore, couples may engage in more or less effective or conscientious assessments of their risk. A few participants described using characteristics-based theories and other "evidence" in determining whether their partners were safe for unprotected sex. This evidence went beyond whether someone "looked healthy" or seemed like "that kind of person." Roberta, a black, HIV-negative 44-year-old, illustrates this point in her comments about her partner, Lionel:

"I really couldn’t tell you [how I knew that he didn’t have a disease]. … He must have been using condoms, because he always had condoms on him."

For Roberta, the fact that Lionel carried condoms meant that he used condoms with other partners and therefore was safe.

Amanda, who had never used condoms with Juan, decided to engage in unprotected sex once she knew that they both had hepatitis C; she took his word that he had tested negative for HIV prior to their meeting. Similarly, Jennifer had decided not to use condoms with her boyfriend, Alberto, upon the realization that they both had hepatitis C infection and had had negative HIV tests before they met.

These examples reveal three approaches to assessing risk with potentially very different levels of success in actually minimizing risk. Roberta's risk assessment relied on characteristics-based theories and Lionel's claim that he used to carry condoms, and involved very little discussion. Amanda's and Jennifer's rested on the fact that both they and their partners had hepatitis C; this information came out over the course of many hours of conversation during which they and their partners discussed their pasts. Finally, José and Linda acknowledged that they had participated in high-risk behaviors before entering drug treatment; they then decided to continue to use condoms while waiting to see if they were still HIV-negative and if their relationship was becoming a committed and monogamous one that included a relatively high level of discussion and negotiation.

While it is clear that several couples consciously engaged in negotiated safety or other strategies to assess or reduce their risk at the beginning of the relationship, their use of these strategies apparently was not consistent throughout the relationship. This was the case even when considerable risk existed, as when one partner was living with HIV or infidelity was suspected.

DISCUSSION

While factors such as substance use, intimate partner violence and self-efficacy play a role in condom use generally, among these 25 couples, these factors seem to have very little effect on the desire or ability to use or insist on condom use within their relationship. Study participants were overwhelmingly marginalized socially and economically, and many abused drugs; nevertheless, they tended to be in long-term, committed relationships. Because of their marginalized and potentially stigmatized status, they may have had a greater need for security, acceptance and reciprocated love than is typical in heterosexual populations at lower risk for HIV infection. Not using condoms, then, stemmed from their desire to establish and maintain a loving and meaningful relationship, at whatever cost, including risk to their health. Most participants were well aware of the risks they were taking, but finding their "soul mate"—a partner who would share and help them navigate through their difficult lives—and achieving a sense of "normalcy" were higher priorities.

Serodiscordant and concordant-positive partnerships are faced with particular challenges in terms of intimacy and the barrier to it that condoms can present. Entangled with the desire and need for physical pleasure are the desire and need to experience physical and emotional intimacy, to express love for one's partner, and to be with that person and fully share his or her experiences, as well as the fear of loss due to death. The challenge for researchers and practitioners is to develop interventions that address these concerns and promote the importance of protection as complementary to the desire for love and intimacy.

Because couples in our study saw condom use as inconsistent with establishing and maintaining a committed, primary relationship, they used other strategies—some more thoroughly than others—to assess and reduce their risk. Several couples described using some form of risk assessment and reduction at the beginning of their relationship, but not consistently throughout the relationship. In fact, many participants ignored or minimized their risk, or acknowledged a real risk of contracting HIV (as in serodiscordant couples), but still chose not to use condoms consistently.

Given the importance of establishing and maintaining committed, romantic relationships, risk assessment and reduction strategies such as negotiated safety must be reconceptualized and communicated to high-risk heterosexuals as a way of increasing intimacy and strengthening relationships. Risk reduction programs should help high-risk individuals understand that ongoing risk assessment, based on open and honest communication, can help couples increase trust and intimacy within their relationship. Similarly, they should explain that periodic testing may be viewed as a way of expressing love and maintaining trust within the relationship. Our finding that many couples were attempting to assess risk even though negotiated safety has not been widely promoted in the United States among heterosexuals suggests that this may be a more acceptable HIV prevention tool for those in committed relationships than might be expected on the basis of earlier research findings.²²

The participation of so many HIV-affected couples in a study about female condoms that did not use HIV status in its inclusion or exclusion criteria may indicate that high-risk couples are interested in safer sex and alternatives to the male condom. Two potential alternatives are the female condom, which is already available (although not always accessible), and microbicides, currently in development. In the part of our study in which these same couples tried the female condom for two weeks, several participants reported use of the method at baseline; early analysis of trial data indicate that couples sometimes preferred the female condom to the male condom and that if the former were more accessible, couples would incorporate it into their repertoire of STD and pregnancy prevention strategies.³⁶ Additionally, as reported previously,³⁷ vaginal microbicides hold promise as an acceptable and desirable option for women drawn from the same community as those in this study.

Limitations

This study has several limitations. For a variety of reasons, our sample was not representative. First, participation was limited to those who were willing to enroll in the study with a partner. Second, while all types of partners (primary, casual, and paying or paid) were eligible, only individuals with primary partners participated, likely because of the in-depth and longitudinal nature of the study. As such, findings are not representative of heterosexuals who are not in serious relationships, or who have casual, paying or paid partners. Third, because of difficulties encountered during recruitment, the sample may have been limited to those with an intrinsic motivation to participate; hence the abundance of couples in which at least one partner was HIV-infected. Fourth, the sample was composed of socially and economically marginalized persons, many of whom abused substances; the findings are therefore not generalizable to individuals with a higher socioeconomic status or without a substance abuse problem. This limitation may also be considered a strength of the study, as perspectives of marginalized populations are often lost in research and the literature. Finally, self-report bias was an inherent risk of the study; however, by interviewing both members of a couple, we were able to validate the consistency of participants’ reports.

Conclusion

A disconnect exists between public health and lay conceptualizations of sex and the nature of sex between "high-risk" individuals. This is evident both in how research questions are asked and in how the resulting data are interpreted. Our findings support the need to explore the context in which sex occurs (i.e., relationships) and suggest that primary, committed relationships may be even more important for this high-risk population than for heterosexual populations at lower risk for contracting HIV.

HIV prevention interventions need to acknowledge the important emotional and social needs that primary relationships satisfy. Greater attention must be paid to developing interventions that acknowledge these needs and do not contradict people's beliefs about their relationships to the point where they reject those interventions as irrelevant.