Correlates of Concurrent Sexual Partnerships Among Young, Rural African American Men

Abstract

Objective

We investigated the social, behavioral, and psychological factors associated with concurrent (i.e., overlapping in time) sexual partnerships among rural African American young men with a primary female partner.

Methods

We recruited 505 men in rural areas of southern Georgia from January 2012 to August 2013 using respondent-driven sampling; 361 reported having a primary female partner and participating only in heterosexual sexual activity. Men provided data on their demographic characteristics and HIV-related risk behaviors, as well as social, behavioral, and psychological risk factors.

Results

Of the 361 men with a primary female partner, 164 (45.4%) reported concurrent sexual partners during the past three months. Among the 164 men with a concurrent sexual partner, 144 (92.9%) reported inconsistent condom use with their primary partners, and 68 (41.5%) reported using condoms inconsistently with their concurrent partners. Having concurrent sexual partnerships was associated with inconsistent condom use, substance use before sex, and self-reported sexually transmitted infections (STIs). Bivariate correlates of concurrent sexual partnerships included incarceration, substance use, early onset of sexual activity, impulsive decision-making, and masculinity attitudes (i.e., men's adherence to culturally defined standards for male behavior). In a multivariate model, both masculinity ideology and impulsive decision-making independently predicted concurrent sexual partnerships independent of other risk factors.

Conclusion

Masculinity attitudes and impulsive decision-making are independent predictors of concurrent sexual partnerships among rural African American men and, consequently, the spread of HIV and other STIs. Developing programs that target masculinity attitudes and self-regulatory skills may help to reduce concurrent sexual partnerships.

African American men who have sex with women (hereinafter referred to as heterosexual) are an understudied group that has experienced rising rates of human immunodeficiency virus (HIV) infection and high rates of other sexually transmitted infections (STIs).¹ Since 2004, HIV diagnoses among African American men who report heterosexual contact have been increasing by more than 9% annually,² and approximately 25% of African American men currently living with HIV/acquired immunodeficiency syndrome (AIDS) reported contracting the disease through heterosexual contact.³ Heterosexual transmission of HIV is facilitated by non-HIV STIs, both inflammatory and ulcerative, which increase HIV infectivity and susceptibility in both women and men.⁴ Thus, risk conferred by STIs acquired in heterosexual relationships affects the spread of HIV in a community and highlights the importance of investigating heterosexual men's behavior.

Engaging in concurrent sexual partnerships has been identified as a potential influence on the HIV/STI epidemics in African American communities.^5–7 Concurrent sexual partnerships describe situations in which an individual has overlapping sexual relationships with more than one person. They can be contrasted with serial monogamy, in which an individual has a sexual relationship with only one partner, with no overlap in time with subsequent partners. Population-based studies have linked concurrent sexual partnerships to male sex, younger age, and African American race.⁷ Structural drivers, such as community poverty and gender ratios, appear to play a key role in the prevalence of African American men's concurrent sexual partnerships.^8,9 In multiethnic samples, associations also have emerged between concurrent sexual partnerships and a range of personal risk factors, including unemployment and economic distress, substance use, history of incarceration, early onset of sexual activity, and perception of partner infidelity.^5,10–12 Less is known, however, about individual differences in the social, behavioral, and psychological factors that predict concurrent sexual partnerships, specifically among African American men.

The African American men in the present study live in small towns and rural communities in southern Georgia. In these communities, interconnected sexual networks increase the risks that concurrent sexual partnerships pose. In addition, this study focused on men with a primary sexual partner. Compared with men without a primary partner, men with a primary partner report more frequent intercourse and less consistency in condom use with their primary partners.^13,14 In the context of concurrent sexual partnerships, these factors amplify exposure to pathogens for men and their sexual partners. Little research, however, focuses on correlates of concurrent sexual partnerships specifically among African American men with a primary partner.

We investigated previously identified personal risk factors, including incarceration history, early onset of sexual activity, and substance use.^6,10,15 We also considered two psychological processes, masculinity attitudes (i.e., men's adherence to culturally defined standards for male behavior) and impulsive decision-making, that have been suggested as targets of study.^15–17 Little research addresses the influence of psychological processes associated with African American men's concurrent sexual partnerships despite the important role psychological factors play in designing individual-, group-, and community-level interventions.

The first psychological risk mechanism is masculinity attitudes. In his anthropological work with African American and Caribbean men, Whitehead¹⁸ described a set of reputation-based attributes that men may adopt to maintain masculine self-esteem. These attributes include sexual prowess, masculine “gamesmanship” skills (e.g., toughness and ability to seduce women), fathering numerous children, and street smarts. In contrast, masculine respectability attributes include marriage, economic provision for one's family, and satisfactory possessions and accomplishments (e.g., a home, higher education, and economic independence). When men from economically disadvantaged backgrounds experience barriers to respect-based pathways to masculinity, they become more likely to express and identify with reputation-based attributes to achieve a sense of masculine self-esteem.^18,19 Masculinity attitudes characterized by endorsement of reputation-based assets are hypothesized to place men at risk for concurrent sexual partnerships.

The second psychological risk mechanism is impulsive decision-making, which has been implicated in a range of sexual risk behaviors, including inconsistent condom use, casual sex, multiple sexual partners, and sex while intoxicated.^20–24 Impulsive behavior has been thought to result from personality factors that include sensation seeking, urgency, and a lack of premeditation and perseverance.²⁵ A number of possible pathways support the link between impulsive decision-making and concurrent sexual partnerships. Impulsive individuals lack the self-regulatory capacity needed to resist hedonistic impulses when opportunities for concurrent sexual activity arise.²² Impulsive decision-making also reduces men's ability to deal with stresses in committed relationships that can undermine sexual fidelity.^24,26,27

METHODS

The African American Men's Health Project (AMP) is a study of the health risk behavior, relationship development, and well-being of young African American men. The sample includes 505 men aged 19–22 years who were recruited from 11 contiguous rural counties in southern Georgia defined as nonurban by the U.S. Census and with population densities of fewer than 100 people per square mile. Eligibility criteria included self-identification as African American, residence in the sampling area, male gender, and age 19–22 years. Data were gathered from January through August 2013. Although AMP focuses on heterosexual men, we did not screen out men who have sex with men because our pilot research indicated few men self-report same-sex activity (<3%).²⁸ The present study focused on 361 men who reported no same-sex activity and reported having a primary partner, defined as “a woman or girl that you have a very special or committed romantic or sexual relationship with, such as a girlfriend or a spouse.”

Participants were recruited using respondent-driven sampling (RDS), a chain-referral protocol designed to reduce biases commonly associated with network-based samples.²⁹ RDS is a preferred method for sampling interconnected but hard-to-reach populations,^30,31 such as young men whose employment and residential situations change frequently.²⁸ Data from a pilot study supported (1) the effectiveness of RDS in obtaining samples that represent African American young adults in rural Georgia counties²⁸ and (2) the use of the RDS post-stratification weighting procedures.

Sampling proceeded as follows: community liaisons (CLs) recruited 54 initial seed participants from the 11 sampled counties. CLs are respected community members who serve as a bridge between participants and our research center. CLs identified young men through their own social networks and described the study to them. Project staff contacted interested men, described the project, determined eligibility, and set up a data collection visit at the participant's home or a convenient community site (usually a private room in the public library). Upon completion of the data collection visits, each of the initial seed participants provided the names of three men in their personal network who met eligibility criteria. Project staff contacted these men regarding participation. As with the seeds, upon completion of data collection, these participants also provided referral information for three network members. For each network member successfully recruited into the study, the referring participant received $25. Self-report data were gathered from participants via audio computer-assisted self-interviews. The user-friendly program guides respondents through the survey; participants with low literacy are assisted through voice and video enhancements. Each participant received $100 at the conclusion of the data collection visit.

Measures

Concurrent sexual activity, other HIV-related risk behaviors, and self-reported STIs.

We assessed concurrent sexual partnerships with the question, “In the past 3 months, have you had sex with another girl or woman while dating your main female partner?” Men also reported how often within the same three-month time frame, during sex with their main partners, they used condoms, engaged in anal sex, or had sex while intoxicated. In each case, the response scale ranged from 1 (never) to 6 (every time). Men also used this response scale to report their use of condoms with their secondary partners. Self-reported STIs were measured using the question, “In the past year, have you been told by a health-care provider that you have a sexually transmitted infection, sometimes called STD or STI? HPV, herpes, HIV, genital warts, syphilis, gonorrhea, and chlamydia are examples of sexually transmitted infections.” Responses to these items were coded to indicate any vs. no instances of risk behavior or self-reported STIs.

Social and behavioral risk factors.

We defined early onset of sexual activity as self-reported age at first intercourse of younger than 14 years. Participants also indicated the number of times during the past three months they had used alcoholic beverages (e.g., beer, wine, wine cooler, or other liquor), had $4 alcoholic drinks at one time, and used marijuana. These items were standardized and summed to form a substance use index (a50.61). This index has been used widely in previous research.^32,33 We assessed past-year incarceration with the item, “In the past year, how much time have you spent in jail or prison?” Responses, which ranged from 0 (none) to 6 (7–12 months), were scored to indicate any vs. no reported incarceration.

Psychological risk factors.

We assessed masculinity attitudes characterized by endorsement of reputation-based attributes using a nine-item scale developed for AMP. The directions read, “Being a ‘man’ means different things to different people. Below are some statements that men may believe show that someone is a ‘real man.’” Each item began with the stem, “A real man …”, and men responded to the items using a scale ranging from 1 (strongly disagree) to 4 (strongly agree). Example items included, “Can handle himself in a fight” and “Has children by many different women.” Cronbach's alpha for the scale was 0.82.

We assessed impulsive decision-making with a brief version of the Urgency, Premeditation, Perseverance, Sensation Seeking, and Positive Urgency, (UPPS-P) Impulsive Behavior Scale.³⁴ The UPPS-P measure assessed men's tendencies toward positive and negative urgency (e.g., “When I am really excited, I tend to get out of control”), a lack of persistence (e.g., “Once I start a project, I almost always finish it,” reverse coded), a lack of premeditation (e.g., “I often get into trouble because I do things without thinking”), and sensation seeking (e.g., “I like to have new and exciting experiences, even if they are illegal”). The items' response scale ranged from 1 (strongly disagree) to 4 (strongly agree). Cronbach's alpha for the total scale was 0.79.

Demographic characteristics.

Participants reported their education levels (coded as <high school diploma, high school diploma/general educational development equivalent, and >high school diploma), current employment status (any employment vs. unemployed), total monthly income, and living arrangements during most of the past year (living with main partner or spouse, living with family of origin, or other [e.g., living alone or with other relatives/friends]).

Plan of analysis.

We performed analyses from June through August 2014. Initial analyses examined the RDS-derived network using the RDS Analysis Tool.³⁵ The statistical theory upon which RDS is based suggests that if peer recruitment proceeds through a sufficiently large number of waves, the composition of the sample will stabilize, becoming independent of the seeds from which recruitment began and thereby overcoming any bias the nonrandom choice of seeds may have introduced. This stable sample composition is termed “equilibrium” and should occur within a modest number of recruitment waves (,4). This analysis provides the information needed to calculate post-stratification weights. We then compared seed participants with those who were part of the networked chains on all study variables. Non-significant differences suggest that the sample may be combined. We then modeled concurrent sexual partnerships as a binary variable in bivariate and multivariate models and used SPSS^® version 22 for analysis.³⁶ We first examined the bivariate associations between concurrent sexual partnerships and demographic variables, HIV-related risk behaviors and self-reported STIs, social and behavioral risk factors, and psychological risk mechanisms. We then conducted a multivariate logistic regression with significant correlates from the bivariate analyses.

RDS analyses on study variables indicated that recruited participants were not biased by the initial seeds' characteristics; sample equilibrium was achieved within two waves of recruitment on all study variables. Results of t-tests comparing, across all study variables, seed participants and participants who were part of networked referral chains were non-significant, indicating the acceptability of combining seeds with recruited participants in the analyses. A post-stratification weight for concurrent sexual partnerships was estimated; however, the adjustment was negligible (population weights of 0.99 for monogamy and 1.01 for concurrent sexual partnerships) and did not affect the study results. Thus, the results we present were derived using raw data.

RESULTS

Table 1 presents the sample demographics and risk factors by concurrent sexual partnership status. Among the 361 AMP participants in a committed relationship, 164 (45.4%) reported concurrent sexual partnerships in the past three months. Among the 164 men who reported concurrent sexual partnerships, 144 (92.9%) reported inconsistent condom use with their primary partners, and 68 (41.5%) reported using condoms inconsistently with their concurrent partners. Bivariate analyses revealed no significant differences in demographic characteristics based on concurrent sexual partnership status. Having concurrent sexual partnerships was associated significantly with self-reported STIs, inconsistent condom use, and substance use before sex. Among social/behavioral risk factors, having concurrent sexual partnerships was associated with early onset of sexual activity, elevated substance use, and incarceration during the past year. Having concurrent sexual partnerships was also associated with reputation-based masculinity attitudes and impulsive decision-making.

Table 1.

Sociodemographic and risk behavior characteristics with differences in concurrent partnership status among rural African American men in South Georgia, January through August 2013

^aBased on chi-square test or t-test of association

SD = standard deviation

GED = general educational development

STI = sexually transmitted infection

HIV = human immunodeficiency virus

NA = not applicable

Table 2 presents the results of the hierarchical logistic regression model. We retained the men's demographic characteristics as control variables in the model and included all risk factors that were significant in bivariate analyses. Masculinity attitudes (odds ratio [OR] 5 2.06, 95% confidence interval [CI] 1.21, 3.50) and impulsive decision-making (OR51.74, 95% CI 1.01, 3.02) were associated with concurrent sexual partnership status independent of other factors. Men who reported reputation-based masculinity attitudes or who reported high levels of impulsive decision-making also reported elevated rates of concurrent sexual partnerships.

Table 2.

Multivariate logistic regression of concurrent sexual partnerships on individual characteristics among rural African American men in South Georgia, January through August 2013

^ap<0.05

^bp<0.01

OR = odds ratio

CI = confidence interval

GED = general educational development

Ref. = reference group

DISCUSSION

Having concurrent sexual partnerships was associated with inconsistent condom use with primary partners, sex while intoxicated, and self-reported STIs in the past year. These men also reported inconsistent use of condoms with their secondary partners. In bivariate analyses, both social/behavioral risk factors (i.e., early onset of sexual activity, incarceration, and substance use) and psychological risk mechanisms (i.e., masculinity attitudes and impulsive decision-making) predicted concurrent sexual partnerships. In multivariate analyses, however, only masculinity attitudes and impulsive decision-making independently predicted concurrent sexual partnerships.

Little research to date has documented concurrent sexual partnership rates among heterosexual African American men in committed relationships. Senn et al.¹⁷ reported that, among men and women in steady relationships who were interviewed at an STI clinic, 64% reported both “steady” and “non-steady” partners. We identified a somewhat lower rate among a sample of young men not selected for high risk. Consistent with other studies, concurrent sexual partnerships predicted STIs³⁷ and covaried with other sexual risk behaviors, including inconsistent condom use with primary partners and substance use prior to sexual activity.^10,38 Of particular concern, only 41.5% of men who engaged in concurrent sex reported using a condom consistently with their other partners.

Among the established social and behavioral risk factors, early onset of sexual behavior, incarceration in the past year, and substance use were associated with concurrent sexual partnerships in bivariate analyses. In multivariate analyses, however, masculinity attitudes and impulsive decision-making were the only significant predictors of concurrency. Young men who endorsed reputation-based attributes as desirable for achieving masculine status were more likely than those who did not to report concurrent sexual partnerships. Several exploratory studies^15–17,39 linked masculinity attitudes to African American men's concurrent sexual partnerships and other high-risk behaviors. The present study confirms the importance of masculinity attitudes and focuses on masculine attributes that have particular relevance to lower-socioeconomic status African American men.^18,40 The robust effects we found for masculinity attitudes suggest that more attention should be given to these beliefs in preventive interventions and media campaigns.

The influence of impulsive decision-making on concurrent sexual partnerships is consistent with studies that link impulsivity to a wide range of risk behaviors, including high-risk sexual activity. Recent studies on impulsive decision-making link high-risk sexual activity with a tendency to discount the value of future rewards vs. immediate ones.^41,42 These studies suggest that some men may prefer the immediate gratification of sexual activity outside of their primary relationships to the potential emotional and health-related rewards of involvement in a monogamous relationship. Importantly, impulsive decision-making is a malleable characteristic that has been targeted successfully in interventions designed to discourage risk behavior by strengthening executive functioning, enhancing the value of delayed rewards, and supporting a closer connection between emotions and self-regulation.^43–46

Limitations

Several limitations of this study are noteworthy. The extent to which the findings are generalizable to African American men in nonrural settings is not clear. Although RDS is a probability sampling method, its effectiveness compared with random sampling for this population is unknown. Also, we focused on men who had a primary partner at the time of data collection; as such, results may not be generalizable to men who are not involved in a steady relationship. Additionally, we measured concurrent sexual activity in the last three months. This time limit likely enhanced participant recall; however, given that concurrent sexual partners may have occurred prior to that time frame, our results may have underestimated the number of men engaged in concurrent sexual activity. Furthermore, the cross-sectional nature of the data limits inferences regarding cause and effect. Finally, unmeasured variables such as relationship factors may play important roles in men's decision to engage in concurrent sexual partnerships

CONCLUSIONS

This study addressed a lack of data on psychological risk mechanisms associated with concurrent sexual partnerships among rural heterosexual African American men. Our findings suggest that programs targeting masculinity attitudes and self-regulatory skills may reduce risk by reducing concurrent sexual partnerships. We also identified important characteristics associated with concurrent sexual partnerships in a group that is not often studied. Because concurrent sexual partnerships and STIs are associated, risk-reduction interventions should address this behavior as a critical component of infection prevention. Concurrent sexual partnerships could be targeted through individual- and small group-level interventions, as well as in neighborhood-level programs. These programs should address masculinity attitudes and impulsive decision-making.