Abstract
This study examined the association of sexual debut and sexual partnering among men in the United States aged 21 to 44 years who participated in the 2006-2010 National Survey of Family Growth. Age at debut was categorized as <15 years, 15 to 17 years, and ≥18 years to permit comparison with previous research. Sexual partnering was defined as being concurrent, serial monogamist, or monogamist in prior year. Eleven percent reported concurrent partnerships and 6% serial monogamy. Sexual debut <15 and 15 to 17 years was associated with concurrency (adjusted odds ratio [aOR] < 15 = 2.22; 95% confidence interval [CI] = 1.37-3.61; aOR 15-17 = 1.69; 95% CI = 1.05-2.74). Irrespective of age at sexual debut, interventions to reduce risky lifetime number of partners may prevent risky sexual behavior in early adulthood and later in life.
Introduction
In the United States, the average age for first vaginal intercourse is 17.1 years among men, with 90% of youth sexually active by age 19 (Kaestle, Halpern, Miller, & Ford, 2005). For people who have sex with members of the opposite sex, early sexual debut is associated with having concurrent sexual partners in adolescence (Kelley, Borawski, Flocke, & Keen, 2003; O’Donnell, O’Donnell, & Stueve, 2001; Ott, Katschke, Tu, & Fortenberry, 2011). The average age of sexual debut has declined since the 1990s for men (Doyle, 2006). Research examining the link between sexual debut and long-term outcomes in men is scant, even though early sexual behaviors may establish a pattern for later ones (Gomez, Speizer, Reynolds, Murray, & Beauvais, 2008; Sandfort, 2008). Early age of sexual debut has been reported to predict a larger number of partners later in life (Kaestle et al., 2005; Kelley et al., 2003) and extramarital sex among men (White, Cleland, & Caral, 2000).
The mechanism through which age at sexual debut in men who have sex with women may influence sexual partnering is not known. There are, however, two plausible explanations. It may well be that timing of first sexual intercourse affects relationship satisfaction (Harden, Quinn, & Tucker-Drob, 2012). Poor satisfaction in relationships may increase cycling through sexual partners. Late timing of first sex may be a marker for an intrapersonal characteristic, such as a secure attachment style (Belsky, Houts, & Fearon, 2010) or strong self-regulatory ability (Moffitt et al., 2011). Thus, timing of first sexual intercourse when one has more fully matured emotionally may provide the opportunity to develop more effective relationship skills than those who experience their first sexual intercourse at a younger age.
The association between age at first sexual intercourse and concurrency and serial monogamy in adulthood remains largely undefined. Using a nationally representative sample, we examined the relationship between timing of first intercourse with a member of the opposite sex and concurrent sexual partnership and serial monogamy in adulthood among men who have sex with women. This research extends the literature in several ways. First, most research to date on age of sexual debut has been conducted among women (Lyons, Giordano, Manning, & Longmore, 2011; Magnusson, Masho, & Lapane, 2012; Rosenthal et al., 2001) despite the fact that men typically have a younger sexual debut (Cavazos-Rehg, 2009) and report higher levels of sexual risk taking (Chandra, Billioux, & Copen, 2012). Second, this article conceptualizes sexual partnering along a three-level continuum: concurrent sexual partnering, serial monogamy, and monogamous relationships. Previous research has dichotomized sexual partnering as concurrent partnerships or not (Adimora, Schoenbach, & Doherty, 2007). This research recognizes that serial monogamy may provide opportunities for “biologic concurrency” if the time period between switching partners is less than the period of infectivity for sexually transmitted diseases (Mercer et al., 2013). Third, while serial monogamy may be one factor associated with transmission of sexually transmitted diseases, research on condom use has been a primary focus to date. This study is novel in that it focuses on relationship patterns. Fourth, we focused on men who have sex with women rather than men who have sex with men because literature on this topic in this population is scant. Last, these data sources available permit the evaluation of the association among men who range from 21 to 44 years of age. Previous studies have focused on adolescence and adulthood before age 30 years. Our analysis will evaluate the partnering behaviors of men further into adulthood.
Our hypothesis was that early age at sexual debut would be associated with concurrent sexual partnerships and serial monogamy in adulthood, but to a lesser extent with the latter. Furthermore, we hypothesized that the impact of early age at sexual debut on sexual partnering in adulthood would be significant, but that its effects would be lessened over time.
Method
Data Source and Sample
The study used data from the continuous 2006-2010 cycle of the National Survey of Family Growth (NSFG). The NSFG collects data on reproductive and family health among men and women aged 15 to 44 years living in households in the United States. The NSFG sampling framework has been described in detail elsewhere (Lepkowski, Mosher, Davis, Groves, & Van Hoewyk, 2010). Briefly, the NSFG identified 110 major areas in the country and sampled from these primary sampling units. The sampling procedures included an oversampling of age and racial/ethnic groups to assure adequate sample size in subgroups. Women professional interviewers were trained to conduct in-person interviews in respondents’ homes from June 2006 to June 2010. Interviews were conducted in English and Spanish. Computer-assisted personal interview (CAPI) and audio computer-assisted self-interviewing (ACASI) were used for sensitive questions about sexual behaviors and drug use. The public use data files released in January 2012 included 10,403 interviews of men (Lepkowski et al., 2010). This research used publicly available de-identified data sets and thus did not require institutional review board approval.
Eligibility Criteria
The sample included men 21 to 44 years of age (n = 7,673). We included men having had at least one woman partner in the previous 12 months because the NSFG questionnaires ask about sexual partnering regarding opposite sex partners (e.g., the variables needed to define sexual concurrent partnerships). The study considered respondents aged 21 years of age or older as adults. There were 2,730 men ineligible because of age and 1,177 ineligible because they did not report a female sexual partner in the previous year. We also excluded men with missing data on key variables (n = 429). The remaining 6,067 men (weighted N = 40,377,309) were eligible for the current study.
Operational Definition of Timing of Sexual Debut
Men were asked, “Have you ever had sexual intercourse with a female (sometimes this is called making love, having sex, or going all the way)?” If yes, they were asked for the month and year of this first intercourse and age at that time. We considered those in the 15 to 17 years age-group to be “average” age of sexual debut, those in the 18 years and above age-group served as the referent group, and those in the ≤15 years of age and below group were considered to have an early debut (Cavazos-Rehg, 2009; Magnusson et al., 2012).
Definition of Monogamy, Serial Monogamy, and Concurrent Sexual Partnerships
Each respondent was asked about the number of partners they had had vaginal sex with in the previous 12 months. For each of the partners reported, the date in months and year of first and last sexual intercourse were asked, and except for any partners identified as currently married to or cohabitating with the respondent, whether or not the partner was “current.” The ACASI interview asked respondents how many sexual partners they had (over the lifetime and in prior 12 months). For up to three discrete opposite sex sexual partners in the past year, the month and year of the first and last vaginal sexual intercourse were reported.
Sexual partnerships were conceptualized in three categories: monogamy, serial monogamy, and concurrency. Monogamy was defined as reporting one sex partner over the course of the previous 12 months. Serial monogamy was defined as more than one sex partner over the past 12 months but with no overlap of first/last sex dates of any other partners. Concurrency was defined as more than one partner in the past 12 months with an overlap of current partner first sex date and previous partner(s) last sex date. Respondents reporting a monogamous relationship in the previous 12 months served as the referent group.
Potential Confounders
We considered the following sociodemographic variables, including current and childhood socioeconomic measures, as potential confounders. Marital status (being unmarried vs. being married), younger age at time of interview, early age when first entering into marriage, long duration of marriage, absence from home, and separation from spouse have been found to be associated with a higher probability of having concurrent partners (Adimora et al., 2007; Adimora, Schoenbach, Taylor, Khan, & Schwartz, 2011). Self-reported race/ethnicity was classified as non-Hispanic White, non-Hispanic African American, Hispanic, or non-Hispanic Other. Age at interview was categorized as 21 to 24 years, 25 to 29 years, 30 to 34 years, 35 to 39 years, or 40 to 44 years. Respondent’s educational attainment was categorized as less than high school, high school graduate, or at least some college. Current household income was evaluated as a percentage of the federal poverty level (FPL; <100% FPL, 100-199% FPL, ≥200% FPL). Relationship status was categorized as never married, currently married or cohabitating, or formerly married. Respondents were asked about the highest level of education that their parents had attained. Parental education level was characterized as “less than high school,” “high school graduate,” or “at least some college or more.”
Analytic Approach
All analyses accounted for the complex sampling design and weighting of the NSFG using SUDAAN (ver. 10.0.1, Research Triangle Institute, NC; Lepkowski et al., 2010). Multinomial logistic regression models (Hosmer & Lemeshow, 2000) were developed to evaluate the association between age at sexual debut and sexual partnering, adjusting for sociodemographic, childhood, and sexual history characteristics. Multinomial logistic regression yields more precision and power than simple dichotomous analysis and allows the comparison of each level of sexual partnering to the referent group one at a time, using separate logistic models for each comparison (Rothman, Greenland, & Lash, 2008). Variables that altered the estimate quantifying the association between age at sexual debut and sexual concurrency by more than 10% were retained. To evaluate the extent to which the impact of age at sexual debut diminishes with time, analyses were stratified by age (categorized as 21-30 years or 31-44 years). Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were derived from these models.
Results
The characteristics of the sample stratified by sexual partnering in adulthood are provided in Table 1. Eleven percent reported concurrent sexual partnerships (n = 919) and 6% serially monogamous relationships (n = 444) in the 12 months prior to interview. Distribution of parental education did not vary by sexual partnering in adulthood. The median number of lifetime partners was 15 among men reporting concurrent sexual partnerships (interquartile range [IQR] = 7-30), 11 among men reporting serial monogamy (IQR = 6-20), and 5 among monogamists (IQR = 2-12). Those in concurrent relationships were more likely to report having had more than two sex partners in the previous year than those practicing serial monogamy (62.6% vs. 26.4%).
Characteristics by Sexual Partnering in the Year Before Interview.
Data are presented as N (weighted percentage).
Age of Debut: Adult Sexual Partnering Relationship
The association between age at sexual debut and sexual partnering in adulthood is provided in Table 2 for the overall sample and stratified by age at interview. Among those reporting age at sexual debut before age 15, 18.8% (n = 312) of men reported concurrent sexual partnerships and 7.9% reported serial monogamy (n = 97). Age at sexual debut was associated with concurrent partnerships in adulthood (aOR<15 versus 18+ = 2.22; 95% CI = 1.37-3.61) and (aOR15-17 versus 18+ = 1.69; 95% CI = 1.05-2.74). The association between age at sexual debut and serial monogamy in adulthood was similar for debut at age <15 years (aOR<15 versus 18+ = 1.52; 95% CI = 0.86-2.67) and sexual debut between 15 and 17 years of age (aOR15-17 versus 18+ = 1.50; 95% CI = 0.95-2.39).
Association Between Age at First Intercourse and Sexual Partnering in Adulthood Overall and Stratified by Age at Interview.
Note. CI = confidence interval. Adjusted for race/ethnicity, marital status, HIV testing history, having a nonmonogamous female partner, ever having been jailed, binge drinking in past year, and use of any drug in past year.
The association between age at debut and concurrency strengthened when stratified by age at interview (Table 2). Among those 21 to 30 years of age at the time of interview, sexual debut earlier than 15 years was associated with concurrency (aOR<15 versus 18+ = 1.88; 95% CI = 1.02-3.44) as was age at sexual debut between 15 and 17 years (aOR15-17 versus 18+ = 1.73; 95% CI = 1.05-2.85) relative to sexual debut ≥18 years of age. For those 31 to 44 years, the association with concurrency for those with a debut under 15 years remained (aOR<15 versus 18+ = 2.65; 95% CI = 1.31-5.36). Relative to age of first intercourse 18+ years, age of debut between 15 and 17 years was associated with serial monogamy (aOR15-17 versus 18+ = 2.17; 95% CI = 1.16-4.05).
Discussion
Overall, this study found that 11% of adult men who have sex with women in the United States reported sexual concurrent partnerships in the year preceding the interview and 6% reported serial monogamy. Age at sexual debut less than 17 years was associated with increased odds of sexual concurrent partnerships in adulthood. The association was present among men aged 21 through 30 years as well as those aged 31 to 44 years of age. Among men aged 21 to 30 years, there was no appreciable difference in odds of sexual concurrency in adulthood for men whose sexual debut was less than 15 years of age and those whose sexual debut was between 15 and 17 years of age. The association between age at sexual debut and serial monogamy was modest and became attenuated when other confounding factors were considered. This extends previous research that has only focused on sexual concurrent partnerships in adolescence or before 30 years of age. Our data support the notion that association between age at sexual debut and sexual partnering extends into middle age as our cohort included men through age 44 years. Our a priori hypothesis was that the strength of the association between age at sexual debut and sexual partnering patterns in adulthood would be lessened with age at time of interview was not supported.
We found the effects of age at sexual debut on sexual partnering behaviors in men who have sex with women linger well into middle age. For youth who are already sexually active, most existing sexual risk reduction programs offer few strategies or approaches beyond return to abstinence, improvement in contraceptive use, or reduction in frequency of intercourse (Manlove, Romano-Papillo, & Ikramullah, 2004). Our findings support the inclusion of messages and skill-building activities focusing on partner reduction, fidelity in romantic relationships, and support continued age-appropriate education for older teens and young adults.
While our preliminary analyses supported a small association between age at sexual debut and serial monogamy, once adjustments for confounders were applied, the association was diminished. Serially monogamous partnerships have the potential for transmission of sexually transmitted diseases owing to “biologic concurrency” if the length of time between sexual partners is less than the infectivity period of the sexually transmitted disease (Brunham & Plummer, 1990; Mercer et al., 2013; Wasserheit & Aral, 1996). Among men practicing serial monogamy, “biologic concurrency” may occur when a person has contracted a sexually transmitted disease from one partner and because of the varying infectivity period and asymptomatic disease, the person may unknowingly transmit a sexually transmitted disease when engaging in sexual relations with a new partner. While technically not a concurrent sexual partnerships, the likelihood of “biologic concurrency” exists when the infectivity period is longer than the time interval between sexual partners. In 2011, the Centers for Disease Control and Prevention estimated that 4,588 new cases of HIV and ~1,600 cases of syphilis occurred among heterosexual men in the United States (Centers for Disease Control and Prevention, 2011). While it is unknown how many of these infections can be attributed to serial monogamy, we believed that studies on serial monogamy are important but lacking in the literature. Because sexual transmission of disease is an ongoing crisis that entails a high toll among men, we thought it prudent to explore factors associated with serial monogamy as the length of time between partners may be less than the infectious period. We do acknowledge, however, that sexual partnering is just one factor contributing to sexually transmitted diseases. Although many studies have focused on condom use, this study specifically has focused on factors associated with sexual partnering.
The practical implications of our findings must be considered. Some existing programs have had a positive impact on the delay of sexual debut among young men (Aten, Siegel, Enaharo, & Aginger, 2002; Jemmott, Jemmott, & Fong, 1998; Lonczak, Abbott, Hawkins, Kosterman, & Catalano, 2002; St. Lawrence et al., 1995) and could be enhanced by our findings. Approaches have included abstinence only programs, comprehensive sex education, HIV/AIDS and other sexually transmitted diseases focused programs, general youth development interventions, and community service learning programs (Manlove et al., 2004). Long term, early interventions are most successful in reducing sexual risk behaviors (Coyle, Kirby, Marin, Gomez, & Gregorich, 2004; Lonczak et al., 2002). Programs offered to youth in middle school and younger could boost the number of young men who delay sexual debut and have long lasting protective effects into later adulthood. Other research indicates programs to delay sexual debut are less successful. In a recent report from the National Campaign to Prevent Teen Pregnancy, more than 20 types of interventions were evaluated and less than half delayed first sex among teens (Manlove et al., 2004). Main reasons for failure of programs included inadequate implementation period (Piper, Moberg, & King, 2000), omission of active skills-building exercises (Aarons, Jenkins, Raine, & El-Khorazaty, 2000), and focusing on age-groups that were older and more likely to be sexually experienced (Manlove et al., 2004). Furthermore, even successful interventions often had short-term effects that would not make a meaningful difference over the course of a man’s lifetime (Jemmott et al., 1998; Miller et al., 1993).
This study has a number of important strengths. This study was large and nationally representative. Oversampling of minority groups provides confidence that sufficient numbers of minorities were included in the analyses. The use of sampling weights helps to account for nonresponse bias and improves the extent to which results can be generalized to the U.S. population. Unlike previous research, this study evaluated serial monogamy in addition to sexual concurrent partnerships. While the nature of sexually transmitted diseases is multifactorial, condom use has typically been an area of research focus. Our study is novel in that it explores serial monogamy as a separate phenomenon. The study of relationship patterns may provide new insights to inform intervention research. The NSFG offered a comprehensive set of variables to adjust for potential confounding. Having such an array of variables is a strength of this study. Last, this study includes men aged through 44 years. Previous research has been limited to adolescence or through age 30 years. That our study is able to evaluate this association into middle age offers a unique perspective on sexual partnering behaviors.
The analyses may have some limitations. The challenges of accurate measurement of sexual behaviors have been acknowledged (Fenton, Johnson, McManus, & Erens, 2001). Our estimate of concurrent sexual partnerships is slightly higher than other reports (Adimora et al., 2007). The slight difference may be due to errors in reports of dates of sexual relationships. Recall error is likely for first sexual experiences, particularly for individuals whose sexual debut occurred long before their interview. In a previous study among young men, while only fair levels of agreement between reported age at first intercourse when interviews were conducted at two times, the difference in age was of only 1 year, with the average divergence in both directions of 1.2 years (Wight & West, 1999). While there may be some bias introduced due to recall of age of sexual debut, especially among older men, the extent of that bias may not be of great concern. All information was self-reported. Since some of the behaviors, particularly sensitive activities related to sexual practices and partnering and substance use, are stigmatized, there is potential for underreporting (Fenton et al., 2001). To limit information bias and underreporting, both CAPI and ACASI were employed. The use of computer-assisted methods to enhance response rates and accuracy, especially using ACASI to limit underreporting when asking sensitive questions, are well supported in the literature (Brewer et al., 2005; Morrison-Beedy, Carey, & Tu, 2006). Finally, overlap of partners was not perfectly measured; in the month, year increments reported, it is possible that partners might not overlap but would be captured in the data as such. This would misclassify more partnerships as “concurrent” than truly were. However, in previous studies comparing the accuracy of similar date comparison methods relatively high agreement was found between the two types of measures (Adimora et al., 2004; Adimora et al., 2011).
In summary, this study expands the literature in several important ways. Much of the previous research on age of sexual debut has been conducted in men. Our study specifically evaluated men as they typically have a younger age of sexual debut (Cavazos-Rehg, 2009) and have higher levels of sexual risk taking (Chandra et al., 2012). We also extend the literature on sexual partnering because we have conceptualized sexual partnering along a continuum that differentiates serial monogamous partnerships from monogamous partnerships. Rather than focus on condom use, we specifically chose to focus on relationship patterns. In addition, much of the research on age of sexual debut is in adolescence. This study extends the literature by evaluating the impact of age at sexual debut on relationship patterns in adulthood. We found that sexual concurrency and serial monogamy were not uncommon in men who have sex with women. We also found that age at sexual debut is associated with sexual concurrent partnerships and that the association remains well into middle age. While beyond the scope of the current study, evaluating the extent to which age at first intercourse influences use of barrier contraceptive methods in adulthood would be an important next step.
Footnotes
Acknowledgements
The authors wish to thank Dr. Christopher Brooks for his editorial comments and support in the preparation of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.