Abstract
Purpose
Epidemic levels of sexually transmitted infections (STIs) among urban youth have drawn attention to the potential role of sex partner selection in creating risk for STIs. The objectives of this study were to describe the ideal preferences and real selection of sex partners, to evaluate sex partner ideal vs. real discordance using quantitative methods and to determine the association between discordance and STI risk behaviors.
Methods
Data are from an urban, household sample of 429 15–24 year olds. Trait clusters were developed for participants’ ratings of their real and ideal sex partners and tested for reliability. Discordance between the ratings of real and ideal partners was measured. Logistic regression was used to assess associations between sex partner discordance and STI risk behaviors.
Results
Real sex partners ratings were often lower than participants’ ideal sex partner ratings. Thirty-three percent of males and 66% of females were discordant on at least one trait cluster. Males discordant on the emotional support they expected of their partner were more likely to report >2 sex partners in the past 90 days (OR=2.13, 95% CI: 1.06, 4.26) and perceived partner concurrency (OR=3.85, 95% CI: 1.53, 9.72). For females, discordance on fidelity or emotional support significantly increased the odds of all risk behaviors.
Conclusions
Male and female adolescents with discordant real and ideal sex partner ratings were more likely to report STI-related risk behaviors. Next steps should involve identification of factors associated with ideal vs. real sex partner discordance such as features of the social context.
Keywords: sex partner selection, STI risk, adolescents, social epidemiology
Introduction
Epidemic levels of sexually transmitted infections (STIs) among inner-city, minority heterosexual adolescents [1,2] have drawn attention to the role that sex partner selection may play in creating risk for STIs among these youth. The overarching hypothesis is that inner-city poor minority young women are more likely to choose sex partners who engage in social behaviors (e.g., drug use, transactional sex) known to increase the risk of STI acquisition and transmission. In order to develop interventions focused on sex partner selection and aimed at reducing morbidity among youth, we must first expand our understanding of how adolescents select sex partners.
Surveys of adults, married couples and college students suggests that individuals select sex partners based on preferences for distinct traits [4–6,10] and that these preferences may differ by gender. When selecting a sex partner men prefer physical attractiveness, while women prefer men with greater financial potential and capacity for emotional intimacy [4,6,10,12].
Little data is available regarding the sex partner selection practices of adolescents, particularly poor, minority adolescents. Qualitative data suggests that inner-city poor minority women are more likely to select sex partners who are at high risk for STIs compared to less poor minority or non-minority young women, not out of preference, but rather because low risk partners are less available in inner-city poor minority neighborhoods.[7,13,15] It is not clear how to directly measure the availability of low risk sex partners. One indirect method would be to compare preferred or “ideal” partners with actual partners. In qualitative interviews, poor urban African American young men and women reliably preferred certain traits in an ideal partner and recognized that the low availability of low risk sex partners resulted in them accepting less than ideal sex partners. For example, female adolescents anticipated accepting a non-monogamous partner although they placed a high value on monogamy in an ideal sex partner. The females reported this was due to their competing desire for a relationship as a respite from economic and familial responsibilities against the backdrop of a male sex partner pool diminished by high rates of male incarceration, morbidity and mortality in their communities [7,13,15]. Sex partner concurrency, having a non-monogamous sex partner, has been found to be associated with risks for STIs in multiple studies including studies among adolescents [16–18].
The objectives of this study were to describe the ideal sex partner preferences and the real sex partner selection patterns, to evaluate sex partner ideal vs. real discordance using quantitative methods and to determine the association between discordance and STI-related risk behaviors among a household sample of urban adolescents and young adults.
Methods
Study Population
A household study was conducted from 2004 to 2007 in Baltimore City among 15–24 year old sexually active Baltimore City residents from selected neighborhoods. The methods for this study have been described [19] however we will review them briefly here.
The sampling strategy was a two stage cluster design with the aim of selecting 65 census block groups and 10 participants per census block group. In the first stage of selection, Baltimore City census block groups (n=710) were selected according to gonorrhea rate for 1994–1999 per 15–39 year olds per 100,000 and an estimation of the number of households within the census block group that were English-speaking and had at least one individual who was 15–24 years old [20]. To be eligible for selection a census block group had to have a gonorrhea rate greater than the 25th percentile and greater than or equal to 35 estimated eligible households. 486 (68%) of census block groups met both of these criteria. These 486 eligible census block groups were ordered by decile of gonorrhea rate, then percent of household incomes below the federal poverty line [20] and finally geographic location in order to ensure that the final census block group sample represented a range of gonorrhea prevalence and poverty and that they were geographically noncontiguous. The final sample of 65 census block groups was selected using stratified probability-proportional-to-size and systematic sampling.
In the second stage of selection, a household sampling frame was created using address lists obtained from three vendors. Two address lists targeted 24 year olds and the third provided a complete list of residential addresses. The sampling frame included 27,194 addresses associated with the 65 census block groups. Next, non-linear optimization was used to select 13,873 households in order to reduce screening costs while controlling for design effects [21].
Procedures
Sample households received a letter describing the study prior to being contacted for enumeration. Enumeration was conducted by telephone when possible (approximately 33%) and otherwise in-person by trained research assistants in order to determine whether an age-eligible individual lived in the household. When more than one age-eligible person lived in a household, one was randomly selected. Further eligibility criteria required that individuals were sexually active, resided at address at which they were contacted and spoke English. Following this screening, informed consent or assent of the parent and/or participant was obtained as appropriate.
Following enrollment, participants completed a privately administered audio computer-assisted self-interview (ACASI). The interview included questions about demographics, sexual history, sexual behaviors and sex partner preferences. Each participant was asked to describe an ideal main sex partner as well as their most recent main sex partner, about whom additional information including partner-specific behaviors and sex partner characteristics were collected. An ideal main sex partner was defined as, “a main sex partner you would like to have, whether this describes your current main partner or not.” A real main sex partner was defined as, “someone that you have sex with, and you consider this person to be the person that you are serious about.” Respondents received gift certificates for their participation. The study protocol was approved by the Johns Hopkins University Institutional Review Board.
Measures
All measures utilized were reported by the participant about themselves and/or their sex partners.
Demographics
Age
Age was calculated using the reported date of birth and the recorded date of visit.
Ethnicity
Ethnicity data were collected by asking the participant to identify their racial background from among Black/African American, White, Asian/Pacific Islander, mixed race or other.
Maternal education level
Maternal education was defined as the highest level of schooling completed by their mother or female guardian. Categories of educational attainment were coded to include less than high school/GED completion, high school/GED completion, some college or completion of college.
Behavioral
Age at first sex
Age at first sex was reported by the participant.
Lifetime number of sex partners
Participants were asked to provide a categorical response to the question, “In your whole life, how many sex partners have you had?” Categorical responses included 1,2,3,4,5–10, 11–20 and greater than 20.
Number of sexual partners in the past 90 days
Number of partners during the last ninety days was reported as a continuous measure.
Index Concurrency
Index, i.e. participant, concurrency was captured by asking about a recent main sex partner in the previous three months followed by the question, “Did you have at least one other partner during the time that you have been or were seeing this partner?”
Sex partner concurrency
Participants were asked, “To the best of your knowledge, did s/he (recent main sex partner) have other sex partners while you two were having a sexual relationship?”
Sex partner traits
For ideal and real main sex partners, participants were asked to rate the same set of seventeen traits.
Ideal main sex partner
Participants were asked, “How important to you is the following quality in an ideal main sex partner?,” using a five-point Likert scale which ranged from, “Not at all important” to “Extremely important”.
Real main sex partner
Each of the 17 traits of interest was included in the statement (“My most recent main sex partner is…”) and participants were asked to what extent each statement described their “most recent main sex partner” using a four-point Likert scale ranging from “Not at all like this” to “Very much like this.”
Statistical Analysis
Analysis was performed separately for males and females due to previous findings of cross-cultural gender differences as well as the hypothesis that within a shared environment adolescents experience gender-specific social pressures resulting in gender-specific sex-partner selection practices.
In order to summarize trait preferences for ideal and real sex partners, trait clusters were created based on our previous qualitative work [7]. Specifically, the 17-item list of partner traits was condensed into five trait clusters: social status (gets respect from people they know, is self-confident, is powerful), financial status (is financially independent, is employed, values educations), physical attractiveness (is sexy, has a nice body, is attractive), fidelity (only has sex with me) and emotional support (makes me feel wanted, can be trusted, tells me about his/her feelings). These 5 trait clusters included 14 of the 17 partner traits. The ratings participants assigned to each trait within a trait cluster were summed to obtain a trait cluster score. The Likert scales used to rank real and ideal sex partners were not equivalent; the scale to rate ideal sex partners had 5 items while the scale for real main sex partners had 4 items. In order to compare real and ideal partners, the 2 most positive response categories for ideal main sex partners were collapsed into one category. Missing scores were imputed if no more than 20 percent of responses to questions regarding ideal main partner traits or real main partner traits were missing. The reliability of the trait clusters was analyzed using Cronbach’s alpha with the exclusion of fidelity which included only one trait. We compared summary values for real and ideal trait cluster scores between and within males and females using paired and unpaired t-tests. In addition, we calculated univariate statistics on selected demographics and sexual risk behaviors to describe participants.
Discordance, defined as a real sex partner trait cluster score that was less than the ideal sex partner trait cluster score, was calculated for each trait cluster and was coded dichotomously .The unadjusted odds ratio for discordance was calculated for the following behaviors previously demonstrated to be associated with STI acquisition: greater than 2 sex partners in the last 90 days, sex partner concurrency and participant concurrency [16–18,22,23].
To determine the odds of selected STI risk behaviors (dependent) associated with discordance (independent), separate logistic regressions were conducted. The selected STI risk behaviors included the participant reporting greater than 2 sex partners in the last 90 days, sex partner concurrency and participant concurrency. Two sensitivity analyses were conducted – one to test whether the results were similar in a study population restricted to heterosexuals and two to test whether the results were similar across age strata (15–18 vs. 19–24 years). All analysis was conducted using Stata (Intercooled Stata 9.0, College Station, TX). Statistical significance was defined as a confidence interval that did not cross one and a p value of less than 0.05.
Results
Study Population
Of the 27,194 addresses in the second stage sampling frame, 50% were fielded and 74% of these were successfully screened. During screening, two of the 65 census block groups were found to be comprised exclusively of retirement communities and thus were excluded. Of enumerated households, 12% had at least one English speaking person between the ages of 15 and 24 of whom 70% completed the eligibility screening with a resulting overall interview response rate of 50.5% [24]. The final study sample included 599 participants residing in 63 census block groups. For this analysis 86 males and 84 males were excluded those because they did not respond to questions about a recent main sex partner and/or failed to answer more than 20% of questions about the traits of either a real or ideal main sex partner. The final sample included 429 participants.
The average age of the participants was 19 (range 15–24) (Table 1). The majority were black (88.3%) and 2.3% reported Hispanic origin. The majority of males and females reported being single (87.7%). 34.5% of males and 30.5% of females reported that their mother had continued their education beyond high school.
Table 1.
Demographic and risk behavior data for participants of household study of 15–24 year-olds in Baltimore, MD, 2004–20071
| Males | Females | |||
|---|---|---|---|---|
| DEMOGRAPHICS | n | % | n | % |
| Gender | 145 | 33.8 | 284 | 66.2 |
| Age (mean, SD) | 142 | 19.0 (2.63) | 275 | 19.5 (2.56) |
| Race | ||||
| Black | 128 | 88.3 | 251 | 88.4 |
| White | 10 | 16.9 | 21 | 7.4 |
| Other | 7 | 4.8 | 12 | 4.2 |
| Maternal Education | ||||
| Less than HS | 32 | 22.1 | 53 | 18.7 |
| High School or GED | 61 | 42.1 | 137 | 48.2 |
| Some College | 21 | 14.5 | 42 | 14.8 |
| College Graduate | 29 | 20.0 | 47 | 16.6 |
| SEXUAL HISTORY | ||||
| Age at sexual debut (mean, SD) | 144 | 13.4 (2.61) | 279 | 14.8 (1.94) |
| Sexual Orientation | ||||
| Heterosexual | 127 | 87.6 | 248 | 87.3 |
| Homosexual or bisexual | 11 | 7.6 | 27 | 9.5 |
| Lifetime # sex partners | ||||
| 1 | 18 | 12.4 | 37 | 13.0 |
| 2 | 11 | 7.6 | 34 | 12.0 |
| 3 | 15 | 10.3 | 35 | 12.3 |
| 4 | 17 | 11.7 | 50 | 17.6 |
| 5–10 | 30 | 20.7 | 66 | 23.2 |
| 11–20 | 23 | 15.9 | 41 | 14.4 |
| >20 | 31 | 21.4 | 16 | 5.6 |
| Number of sex partners in the past 90 days (median) | 144 | 1 | 284 | 1 |
| History of gonorrhea | 16 | 11.0 | 42 | 14.8 |
| History of Chlamydia | 19 | 13.1 | 82 | 28.9 |
| Index concurrency | 61 | 42.1 | 68 | 23.9 |
| Sex partner concurrency | 24 | 16.6 | 68 | 23.9 |
n’s and percentages unless otherwise noted
Regarding history of sexual activity, mean age at sexual debut was 13.4 years and 14.8 years for males and females respectively. The majority of male and female participants were heterosexual (87.6% and 87.3%, respectively). There was considerable variability in the number of lifetime sex partners. 80% of males and 75% of females reported more than 2 lifetime sex partners. A history of chlamydia was reported by 28.9% of females and 13.1% of males while 14.8% and 11.0% of females and males respectively reported a history of gonorrhea.
Trait Clusters
The trait clusters demonstrated good to very good reliability for real main partners for both males and females using Cronbach’s alpha (range 0.71–0.91; Table 2). For females, the trait clusters of emotional support, social status and financial status should be interpreted with some caution given borderline Cronbach’s alpha results in the case of ideal main partners.
Table 2.
Real and ideal sex partner traits of interest combined into trait clusters with associated alphas in household study of 15–24 year-olds in Baltimore, MD, 2004–2007
| Cronbach’s Alpha | ||||||
|---|---|---|---|---|---|---|
| Ideal Main Partner | Real Main Partner | |||||
| Trait Cluster | # Traits | Traits | Male | Female | Male | Female |
| Emotional Support | 3 | Makes me feel wanted | ||||
| Trustworthy | ||||||
| Discusses feelings | 0.85 | 0.61 | 0.74 | 0.80 | ||
| Social Status | 3 | Is respected | ||||
| Is self-confident | ||||||
| Is powerful | 0.78 | 0.54 | 0.91 | 0.71 | ||
| Financial Status | 3 | Is financially independent | ||||
| Is employed | ||||||
| Values education | 0.73 | 0.65 | 0.79 | 0.71 | ||
| Physical Attractiveness | 3 | Is sexy | ||||
| Has a nice body | ||||||
| Is attractive | 0.88 | 0.86 | 0.91 | 0.84 | ||
| Fidelity | 1 | Only has sex with me | ||||
Comparing ideal main and real main sex partners separately within males and females, the negative difference in trait cluster scores suggests that recent real main sex partners did not meet the expectations of ideal main sex partners. This difference was significant for emotional support and financial independence among males and for emotional support, financial independence and fidelity among females (Table 3). In contrast, both males and females rated the physical attractiveness of their real main sex partner more highly than that of an ideal main sex partner, although the difference was only significant for females.
Table 3.
Comparison of trait cluster scores for real and ideal sexual partners by gender in household study of 15–24 year-olds in Baltimore, MD, 2004–2007
| Within Gender | Between Gender | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Male1 | Female1 | Ideal Main2 | Real Main2 | |||||||||
| Males vs. Females | Males vs. Females | |||||||||||
| Ideal Main | Real Main | Δ | P value | Ideal Main | Real Main | Δ | P value | Δ | P value | Δ | P value | |
| Emotional support | 10.93 | 10.4 | −0.53 | 0.02 | 11.5 | 10.6 | −0.96 | 0.0000 | −0.64 | 0.0000 | −0.22 | 0.3 |
| Social status | 10.1 | 9.7 | −0.38 | 0.2 | 10.7 | 10.5 | −0.14 | 0.3 | −0.57 | 0.0008 | −0.81 | 0.0003 |
| Financial independence | 9.4 | 8.9 | −0.57 | 0.04 | 10.4 | 10.0 | −0.42 | 0.01 | −1.00 | 0.0000 | −1.14 | 0.0000 |
| Physical attractiveness | 10.3 | 10.4 | 0.07 | 0.8 | 8.6 | 10.1 | 1.50 | 0.0000 | 1.69 | 0.0000 | 0.26 | 0.3 |
| Fidelity | 3.57 | 3.48 | −0.09 | 0.4 | 3.79 | 3.48 | −0.31 | 0.0000 | −0.20 | 0.001 | −0.01 | 1.0 |
paired t-test;
unpaired t-test;
maximum score = 12 except for fidelity (max score = 4)
Comparing the trait cluster scores between genders, there were differences reported for ideal main sex partners and real main sex partners (Table 3). For ideal main sex partners, males as compared to females were significantly more likely to rate physical attractiveness (difference 1.69, p value < 0.00) as important whereas females as compared to males were significantly more likely to rate financial independence (difference −1.00, p value < 0.00) as important. For real main sex partners, females as compared to males were significantly more likely to rate financial independence (difference −1.14, p value < 0.00) and social status (difference −0.81, p value < 0.00) as important.
Thirty-three percent of male participants and 66% of female participants were discordant on at least one trait cluster. Discordance was associated with STI-related risk behaviors (Table 4). The association was consistent across trait clusters, but more pronounced for fidelity and emotional support. Males who reported emotional support discordance were more likely to report greater than 2 sex partners in the past 90 days (OR=2.13, 95% CI: 1.06, 4.26) and sex partner concurrency (OR=3.85, 95% CI: 1.53, 9.72).
Table 4.
Unadjusted odds ratio (OR) of sexual risk behaviors according to trait cluster discordance by gender in household study of 15–24 year-olds in Baltimore, MD, 2004–2007
| Females | ||||||
|---|---|---|---|---|---|---|
| >2 sex partners in the past 90 days | Index concurrency | Sex partner concurrency | ||||
| OR | 95% CI | OR | 95% CI | OR | 95% CI | |
| Emotional Support | 2.17 | 1.11, 4.24 | 2.23 | 1.28, 3.88 | 3.27 | 1.85, 5.79 |
| Social Status | 1.82 | 0.92, 3.60 | 1.41 | 0.79, 2.52 | 1.18 | 0.65, 2.14 |
| Financial Status | 1.46 | 0.75, 2.83 | 0.95 | 0.54, 1.66 | 1.44 | 0.83, 2.52 |
| Physical Attractiveness | 1.52 | 0.73, 3.18 | 1.46 | 0.79, 2.73 | 1.26 | 0.67, 2.39 |
| Fidelity | 3.10 | 1.55, 6.19 | 2.18 | 1.20, 3.97 | 6.68 | 3.52, 12.67 |
| Males | ||||||
| >2 sex partners in the past 90 days | Index concurrency | Sex partner concurrency | ||||
| OR | 95% CI | OR | 95% CI | OR | 95% CI | |
| Emotional Support | 2.13 | 1.06, 4.26 | 1.15 | 0.58, 2.27 | 3.85 | 1.53, 9.72 |
| Social Status | 2.13 | 1.06, 4.26 | 1.29 | 0.65, 2.56 | 1.86 | 0.76, 4.56 |
| Financial Status | 1.14 | 0.58, 2.24 | 0.51 | 0.26, 1.01 | 1.67 | 0.68, 4.06 |
| Physical Attractiveness | 1.88 | 0.90, 3.92 | 1.54 | 0.74, 3.20 | 2.89 | 1.15, 7.28 |
| Fidelity | 2.47 | 1.17, 5.23 | 1.14 | 0.54, 2.41 | 2.05 | 0.80, 5.27 |
For females, discordance on either fidelity or emotional support significantly increased the odds of having greater than 2 sex partners in the last 90 days, and reporting sex partner concurrency and index concurrency. The strongest association was between discordance on fidelity and sex partner concurrency (OR=6.68, 95% CI: 3.52, 12.67).
For both males and females, the findings of trait cluster discordance were unchanged when the sample was limited to participants who self-reported being heterosexual or when the sample was stratified by age (15–18 vs. 19–24 years).
Discussion
Among this household sample of urban adolescents and young adults, we were able to confirm and show adequate to very good reliability of a set of trait clusters which qualitative data suggest play an important role in sex partner selection. Comparing trait clusters for ideal vs. real main sex partners within males and females showed that recent real main sex partners did not meet the expectations of ideal main sex partners on all trait clusters (except physical attractiveness) although only significant in some. Comparing trait clusters between genders for ideal main sex partners, males rated physical attractiveness as important whereas females rated financial independence as important. For real main sex partners, females were significantly more likely than males to rate financial independence and social status as important. These findings are similar to previous findings among adults, married couples and college students which suggests that sex partners are selected based on preferences for distinct traits [4–6,10], preferences differ by gender and males tend to prefer physical attractiveness, while females prefer males with greater financial potential [4,6,10,12]. We also found that ideal sex partner traits were often discordant with real sex partner traits. These findings are similar to our qualitative findings among a similar population. Furthermore, males and females who were discordant on the emotional support they received from their partners were significantly more likely to report important STI-related risk behaviors including having had greater than 2 partners in the past 90 days, sex partner concurrency and index concurrency. Additionally, females, but not males, discordant on emotional support were more likely to report their own (i.e. index) concurrency and females discordant on fidelity were more likely to report all STI-risk behaviors. These findings are a first step in quantifying discordance and understanding its impact on STI risk.
An important limitation of this study is its cross-sectional design. Whether and to what extent adolescents select discordant sex partners over time would be valuable information. However, these data come from a large, rigorously conducted household sample and provide a unique opportunity to analyze discordance in ideal and real sex partners among adolescents. Additional limitations are the reliance on self-report of STI risk behaviors and egocentric reporting, i.e. participants reported on the behaviors of their sex partners in the case of concurrency. Finally, for questions regarding the traits of an ideal main sex partner, the gender of that ideal main partner was not asked or specified. This precludes an analysis of the role of the gender of the participant or their partner on discordance between real and ideal main partners.
Sex partner selection is an important component of sexual behavior for all adolescents and may be an important determinant of risk for STIs among inner-city poor minority females in particular. This study suggests that adolescents express specific preferences in ideal partners and when adolescents report that specific preferences are not met in a real sex partner, they are more likely to report STI risk behavior. Our objective was to use quantitative methods to test compelling theories generated by qualitative studies. While it is premature to use these findings to recommend a specific public health intervention, when reviewed in conjunction with previous qualitative work, these results can be used to advocate for further investigations of factors influencing adolescent sex partner selection, particularly features of the social context.
Acknowledgments
This work was supported by a grant from the National Institute of Child Health and Human Development (R21HD052438) awarded to J.M. Ellen as well as grant from the National Institute on Drug Abuse (K01DA022298) awarded to J.M. Jennings.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Sexually transmitted diseases surveillance. Atlanta, GA: Centers for Disease Control and Prevention; 2007. [Accessed June 2, 2010]. Online. Available from: http://www.cdc.gov/std/stats07/toc.htm. [Google Scholar]
- 2.African americans and HIV/AIDS [Online] Menlo Park, CA: Henry J. Kaiser Family Foundation; [Accessed June 2, 2010.]. Available from: http://www.kff.org/hivaids/upload/African-Americans-and-HIV-AIDS-Fact-Sheet.pdf. [Google Scholar]
- 3.Botwin MD, Buss DM, Shackelford TK. Personality and mate preferences: Five factors in mate selection and marital satisfaction. J Pers. 1997;65(1):106–36. doi: 10.1111/j.1467-6494.1997.tb00531.x. [DOI] [PubMed] [Google Scholar]
- 4.Buss DM, Abbott M, Angleitner A, et al. International preferences in selecting mates: A study of 37 cultures. Journal of Cross-Cultural Psychology. 1990;21(1):5–47. [Google Scholar]
- 5.Buss DM, Shackelford TK, Kirkpatrick LA, et al. A half century of mate preferences: The cultural evolution of values. Journal of Marriage and Family. 2001;63(2):491–503. [Google Scholar]
- 6.Lippa RA. The preferred traits of mates in a cross-national study of heterosexual and homosexual men and women: An examination of biological and cultural influences. Arch Sex Behav. 2007;36(2):193–208. doi: 10.1007/s10508-006-9151-2. [DOI] [PubMed] [Google Scholar]
- 7.Andrinopoulos K, Kerrigan D, Ellen JM. Understanding sex partner selection from the perspective of inner-city black adolescents. Perspectives on Sexual and Reproductive Health. 2006;38(3):132–8. doi: 10.1363/psrh.38.132.06. [DOI] [PubMed] [Google Scholar]
- 8.Anderson E. Sex codes and family life among poor inner-city youths. Ann Am Acad Pol Soc Sci. 1989;501:59–78. [Google Scholar]
- 9.Kerrigan D, Andrinopoulos K, Johnson R, et al. Staying strong: Gender ideologies among African-American adolescents and the implications for HIV/STI prevention. J Sex Res. 2007;44(2):172–80. doi: 10.1080/00224490701263785. [DOI] [PubMed] [Google Scholar]
- 10.Feingold A. Gender differences in effects of physical attractiveness on romantic attraction: A comparison across five research paradigms. J Pers Soc Psychol. 1990;59(5):981–93. [Google Scholar]
- 11.Geary DC, Vigil J, Byrd-Craven J. Evolution of human mate choice. J Sex Res. 2004;41(1):27–42. doi: 10.1080/00224490409552211. [DOI] [PubMed] [Google Scholar]
- 12.Li NP, Kenrick DT, Bailey JM, et al. The necessities and luxuries of mate preferences: Testing the tradeoffs. J Pers Soc Psychol. 2002;82(6):947–55. [PubMed] [Google Scholar]
- 13.Sobo EJ. Inner-city women and AIDS: The psycho-social benefits of unsafe sex. Cult Med Psychiatry. 1993;17(4):455–85. doi: 10.1007/BF01379310. [DOI] [PubMed] [Google Scholar]
- 14.Ferguson YO, Quinn SC, Eng E, et al. The gender ratio imbalance and its relationship to risk of HIV/AIDS among african american women at historically black colleges and universities. AIDS Care. 2006 May;18(4):323–31. doi: 10.1080/09540120500162122. [DOI] [PubMed] [Google Scholar]
- 15.Adimora AA, Schoenbach VJ, Martinson FE, et al. Social context of sexual relationships among rural african americans. Sex Transm Dis. 2001 Feb;28(2):69–76. doi: 10.1097/00007435-200102000-00002. [DOI] [PubMed] [Google Scholar]
- 16.Potterat JJ, Zimmerman-Rogers H, Muth SQ, et al. Chlamydia transmission: Concurrency, reproduction number, and the epidemic trajectory. Am J Epidemiol. 1999;150(12):1331–9. doi: 10.1093/oxfordjournals.aje.a009965. [DOI] [PubMed] [Google Scholar]
- 17.Jennings J, Glass B, Parham P, et al. Sex partner concurrency, geographic context, and adolescent sexually transmitted infections. Sex Transm Dis. 2004;31(12):733–9. doi: 10.1097/01.olq.0000145850.12858.87. [DOI] [PubMed] [Google Scholar]
- 18.Rosenberg MD, Gurvey JE, Adler N, et al. Concurrent sex partners and risk for sexually transmitted diseases among adolescents. Sex Transm Dis. 1999;26(4):208–12. doi: 10.1097/00007435-199904000-00004. [DOI] [PubMed] [Google Scholar]
- 19.Jennings JM, Taylor R, Iannacchione VG, et al. The available pool of sex partners and risk for a current bacterial sexually transmitted infection (STI) Ann Epidemiol. doi: 10.1016/j.annepidem.2010.03.016. In Press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Census 2000 summary file 3 technical documentation. U.S. Census Bureau; 2002. [Accessed June 2, 2010]. [Online] Available from: http://www.census.gov/prod/cen2000/doc/sf3.pdf. [Google Scholar]
- 21.Chong EKP, Zak SH. An introduction to optimization. New York, NY: John Wiley & Sons; 1996. [Google Scholar]
- 22.Vivancos R, Abubakar I, Hunter PR. Sex, drugs and sexually transmitted infections in british university students. International Journal of STD and AIDS. 2008;19(6):370–7. doi: 10.1258/ijsa.2007.007176. [DOI] [PubMed] [Google Scholar]
- 23.Adimora AA, Schoenbach VJ, Doherty IA. HIV and african americans in the southern United States: Sexual networks and social context. Sex Transm Dis. 2006;33(7S):S39–45. doi: 10.1097/01.olq.0000228298.07826.68. [DOI] [PubMed] [Google Scholar]
- 24.American Association for Public Opinion Research (AAPOR) Standard definitions: Final dispositions of case codes and outcome rates for surveys. 4. Lenexa, Kansas: American Association for Public Opinion Research (AAPOR); 2006. [Google Scholar]