Despite longstanding knowledge that there are 3 dimensions of sexual orientation—identity, behavior, and attraction—that are not always concordant in individuals,1,2 women who engage in same-sex sexuality without espousing a sexual minority identity (e.g., lesbian or bisexual) have rarely been studied. Rather, health research on those with apparent discordance among sexual orientation dimensions has focused primarily on HIV risk in heterosexual- or nongay-identified men who have sex with men (MSM), because of concerns about risk to female partners,3,4 low rates of condom use and HIV testing,5,6 alcohol and drug use in sexual contexts,7 and lack of access to HIV prevention information in gay communities.3,6,8
Discordant sexual orientation identity and behavior in women has been of lesser interest because sexual contact between women is not considered a priority for research on sexually transmitted infections (STIs), despite recent evidence of risk.9,10 Even outside of STI-related research, health, and sociodemographic information about heterosexual women with female sex partners is virtually nonexistent.
A number of studies in developmental psychology have explored sexual orientation development in “mostly heterosexual” women or women with nonexclusive attractions.11–13 Additionally, epidemiologic studies have examined discordance between behavior and identity in women, but have not related it directly to specific health outcomes.2,14 A search for health outcomes research on heterosexual women with female sex partners located only six published papers. These papers explored: physical health outcomes15; preventive healthcare16; substance use and coerced sexual activity in young, low-income women17; varied health outcomes primarily to assess methodological issues18; and alcohol use and drinking contexts.19–20 Most physical and mental health literature on sexual minority women explored health outcomes related to sexual minority identities or alternately behavior-based groups of all women who indicated sex with women. Only recently have health researchers begun to address the complexity of sexual orientation in women by collecting and using information on both identity and behavior. However, many studies combined these 2 dimensions into a single composite measure of sexual minority status21–23 or conducted parallel analyses using the 2 measures,24 without analyzing specific identity-behavior groups.
Previous research has suggested higher prevalences of self-reported genital herpes and tobacco use in heterosexual women with female partners18 and high rates of coerced sexual behavior and substance use among young, low-income heterosexual women with both male and female partners.17 Results for alcohol use have been mixed.17–20 The current study was undertaken to build on these preliminary findings. The primary objective was to compare heterosexual women who reported female sex partners to exclusively heterosexual women, and bisexual- and homosexual-identified women on demographics, substance use, sexual health and risk behaviors. Understanding the relationship between same-sex behavior in heterosexual women and sexual risk and substance use has implications for HIV and STI prevention, and for efforts in addressing substance use and health services use. Moreover, it will add to current knowledge of the complex relationship between sexual orientation identity, behavior and attraction.
MATERIALS AND METHODS
Data and Study Sample
Data from the 2002 US National Survey of Family Growth (NSFG) were used to compare sexual risk behaviors and substance use for heterosexual-identified women who reported female sex partners either in the past year or more remotely, exclusively heterosexual women, and homosexual and bisexual women. The 2002 NSFG is a national probability-based household survey of 7643 female and 4928 male participants aged 15 to 44 years in which data were gathered on family life, marriage, pregnancy, infertility, contraception, and sexual and reproductive health. Participants were selected using multistage area probability sampling. Survey data were collected by in-home interviews in English or Spanish, with sensitive items collected using audio computer-assisted self interviews (ACASI). Signed consent forms were obtained at the time of interview. NSFG methodology is well documented,25 and information and selected data are available on the US National Center for Health Statistics website at http://www.cdc.gov/nchs/nsfg.htm.
The overall female response rate was 80%, and a total of 6493 women aged 20 to 44 completed the survey. Of these, 6356 (98%) completed the ACASI portion of the survey, which included information about sexual orientation and sex partner history. The 269 women (4.2%) who reported no prior male or female sex partners were excluded, as no information existed with which to classify the sex of their sex partners, resulting in a final sample size of 6087. For comparative analyses of sexual health and substance use, 243 participants who identified their sexual orientation as “something else” were omitted, as they likely represent a heterogeneous group of those who have alternate sexual minority identities, did not understand the question, or found it irrelevant or inappropriate.18 Thus, it was not possible to ascertain whether their orientation identity was concordant or discordant with same-sex sexual behavior.
NSFG protocols and materials were approved by the CDC/NCHS Research Ethics Review Board. Data analysis protocols for the current analyses were declared exempt by the Research Ethics Board at The University of Western Ontario.
For sexual orientation identity, participants indicated whether they identified as “heterosexual, homosexual, bisexual, or something else.” In earlier analyses, it was noted that 25% of participants who indicated “homosexual” identity had never had a female sex partner, had indicated they were “only attracted to males,” and had ≥1 male sex partner in the past year.18 These participants appeared to be misclassified and were excluded from the current analysis. There was no apparent misclassification in the other direction—no participants who indicated a “heterosexual” identity, but only female sex partners and attraction only to females.
From these initial NSFG categories, 5 identity-behavior groups were created as a basis for all comparisons. Homosexual- and bisexual-identified women formed the first 2 comparison groups. The remaining three consisted of heterosexual-identified women, subdivided according to their history of female sex partners. Heterosexual participants were coded as having a recent female sex partner whether they indicated ≥1 female sex partner in the past 12 months and as having a nonrecent female sex partner whether they indicated ≥1 female sex partner in their lifetime, but none in the past 12 months. Heterosexual-identified women who indicated no female sex partner and ≥1 male sex partner in their lifetimes—with sex with males defined as oral, anal, or vaginal—were coded as exclusively heterosexual women.
All measures of substance use were assessed over the past 12 months. Tobacco use was coded as smoking daily versus less than daily or not at all. Binge drinking was defined as having 5 or more drinks in a couple of hours “about once a month” or more, and weekly drinking as consuming alcohol “about once a week” or more often. Use of marijuana and cocaine (powder or crack cocaine) were coded as any use during the past 12 months versus no use.
A range of sexual behaviors and reproductive indicators were chosen for comparison. Participants were compared on self-report of ever having been pregnant, having been diagnosed with genital herpes, and ever having had anal sex with a male partner. Past-year behavioral risk measures included being paid for sex, having a bisexual male partner, having a nonmonogamous male partner, and having sex with a male partner while high. Protective behaviors assessed include past-year Pap tests, past-year STI testing, and condom use with male partners during last vaginal sex. Participants also indicated their age at first vaginal sex with a male partner, and their lifetime numbers of male and female sex partners. Because lifetime number of sex partners was highly skewed for both male and female partners, dichotomous variables were created representing 8 or more male partners, and 3 or more female partners. These cut-points were based on the weighted 75th percentiles among women who had ≥1 partner of that sex.
Weighted frequencies for sexual orientation identity-behavior groups were calculated, and weights within each group were summed to estimate the number of individuals in these groups within the United States. For each identity-behavior group, frequencies were calculated to describe sexual attraction ranging from “only attracted to males” to “only attracted to females.”
Demographics, substance use, and sexual behavior variables were described using weighted means and frequencies with associated 95% confidence intervals. Weighted medians were calculated for lifetime male and female sex partner numbers. All analyses were conducted using survey procedures in SAS 9.2, and were weighted to represent population estimates for US women aged 20 to 44 years. Means were calculated using PROC SURVEYMEANS, and comparisons of means across identity-behavior groups using ANOVA tests in PROC SURVEYREG. Frequencies were calculated using PROC SURVEYFREQ with Rao-Scott chi-square tests for differences across groups. Where tests indicated significant differences overall, pairwise tests were conducted with heterosexual women who had a past-year female sex partner as the reference group.
To determine to what extent observed differences may be due to demographic differences between groups, a second set of adjusted comparisons was made using logistic regression (PROC SURVEYLOGISTIC) or linear regression (PROC SURVEYREG). Comparisons by sexual orientation identity-behavior group were adjusted for age, marital/relationship status, and ethnoracial group. For this analysis, a continuous measure of age was used. Ethnoracial group was dummy coded into 4 categories: Hispanic, white non-Hispanic, black non-Hispanic, and other. Marital/relationship status was similarly coded as a series of dummy variables representing: married; living with a partner; divorced, separated or widowed; and never married. All statistical analyses account for the complex sampling method, and use Taylor series linearization to calculate variances.
Both a heterosexual identity and a history of sex with at least 1 female partner were indicated by an estimated 7.9% of US women aged 20 to 44, a figure approximately double the estimated 3.6% of women who identified as homosexual (1.0%) or bisexual (2.6%). Given the options of “homosexual,” “bisexual” or “heterosexual,” 3.5% of US women opted for “something else.” Among heterosexual women who had ever had a female sex partner, 20% had one within the past year. Thus, an estimated 3,811,501 heterosexual-identified US women aged 20 to 44 have ever had a female sex partner, 755,622 within the past year.
Sexual attraction to male and female partners is presented in Figure 1. About half of heterosexual-identified women with a recent or nonrecent female sex partner indicated attraction to “only males” and nearly all of the remainder to “mostly males.” Whereas these women had less exclusive sexual attraction than heterosexual women with only male partners—93% of whom were “only attracted to males”—they did not demonstrate the breadth of attraction across the spectrum indicated by bisexual women or the female-focused attraction indicated by homosexual women.
Demographic comparisons are presented in Table 1. Sexual orientation identity-behavior groups varied significantly on age, ethnoracial group, and marital status. Heterosexual women with a past-year female sex partner had the lowest mean age (28.7 years) and were unlikely to be currently married. Differences in ethnoracial distribution across identity-behavior groups resulted primarily from differences between Hispanic and non-Hispanic women. When race and Hispanic ethnicity were analyzed separately (results not shown), Hispanic ethnicity was strongly associated with identity-behavior group (P < 0.0001), but race was not (P = 0.35). Compared to non-Hispanics, Hispanic women were less likely to identify as “homosexual.”
Comparisons of substance use across orientation identity-behavior groups are presented in Table 2. Bivariate comparisons across all groups were significant for each of the 5 substance measures, with heterosexual women who had a past-year female sex partner having the highest point estimates of prevalence of use for each measure. In pairwise comparisons with their exclusively heterosexual peers, they were significantly more likely in both bivariate and adjusted analysis to use each substance type. They were about twice as likely to smoke tobacco (46% vs. 19%) or drink alcohol regularly (49% vs. 24%), and three times as likely to binge drink (34% vs. 11%). Over the past year they were 5 times as likely to use marijuana (58% vs. 11%), and 9 times as likely to use cocaine (19% vs. 2%). Heterosexual women with a past-year female partner also had higher prevalences than homosexual women on all measures but weekly alcohol use, though they did not differ significantly on binge drinking in adjusted analysis. They differed significantly from bisexual women only on marijuana use, for which they had significantly higher past-year use than all other groups. Adjustment for age, marital/relationship status, and ethnoracial group resulted in loss of significance for 3 of 4 statistically significant bivariate comparisons between heterosexual women with recent and more remote female sex partners. Only 1 of 10 significant comparisons with other groups similarly lost significance on adjustment.
Sexual health and reproductive measures are presented in Table 3. Most heterosexual women with a past-year female partner had only one such partner in their lifetime, and 98% had a past-year male partner. Heterosexual-identified women who had a same-sex partner, either in the past year or more remotely, were significantly more likely to have high numbers of male sex partners, even after adjusting for demographic variables, and were significantly less likely than homosexual or bisexual women to have high numbers of female sex partners. Median lifetime male partners were 10 for heterosexual women with a recent or nonrecent female sex partner, 4 for exclusively heterosexual women, 1 for homosexual women, and 7 for bisexuals. Median lifetime female partners were 1 for heterosexual women with a recent or nonrecent female sex partner, 0 (by definition) for exclusively heterosexual women, 4 for homosexual women, and 2 for bisexual women. All sexual and reproductive measures other than receiving a Pap test within the past year differed across orientation identity-behavior groups. Similar to heterosexual women with less-recent female partners and to bisexuals, but at a significantly higher frequency than for homosexual and exclusively heterosexual women, 66% of heterosexual women with a past-year female partner had ever had anal sex with a male partner. About 5% had a bisexual male partner in the past year. These differences remained significant in adjusted analyses.
Whereas sexual health results for heterosexual women with a past-year female sex partner were most similar to heterosexual women with a female partner in the more remote past and to bisexual women, there were distinct differences. They were significantly more likely than any other group to have had a nonmonogamous male partner (40%) and to have engaged in sex while high (69%) during the past year, both in bivariate and adjusted analyses. Moreover, their first vaginal intercourse occurred at a younger age (mean = 15.1 year) than all groups other than bisexuals. Reflecting protective behavior appropriate to these risks, they were also more likely than other groups, with the exception of heterosexual women with more remote female-female sexual behavior, to report having been tested for sexually transmitted infections during the past year. However, they were no more likely to report condom use at last vaginal sex.
In contrast with analysis of substance use variables, adjustment for demographic variables affected few associations between identity-behavior group and sexual health measures. Where changes in significance occurred, P values were close to the 0.05 cut-off for significance.
Most studies of female same-sex sexual behaviors have not included many heterosexual women, or have excluded them outright. In contrast, this population-based study provided data on sexual orientation identity, attraction and behavior for a probability-based household sample of US women, and allowed for more detailed study of sexual health and substance use, and deeper insights into the complex relationships between dimensions of sexual orientation. Despite this strength, this study also has some limitations. Sexual orientation, sexual behavior and substance use data are of necessity self-reported. Whereas ACASI provides the most sensitive mode for collecting self-reported data on sensitive topics, underreporting remains a possibility. The stratified probability design of the survey was designed to allow results to be generalized to US women aged 20 to 44 who are living in a noninstitutional setting. Results must be interpreted in the context of this specific population and would not apply, for example, to women in prisons or the military, or to older women.
Heterosexual-identified women with a recent or more remote female sex partner indicated attraction levels consistent with their identity: 99% were attracted only or mostly to males. The extent to which this may be due to social desirability resulting in unwillingness to admit higher levels of attraction to females in a survey, or even to oneself, is unknown. For some women, sex with other women may be highly situationally specific, and not the result of a strong and generalized attraction to women.12
Heterosexual women with female sex partners differed from their exclusively heterosexual peers not only in their sexual partnerships, but also in their substance use, sexual risk taking, and sexually transmitted infection testing behaviors. Whereas previous research found that “mostly heterosexual” women held a unique position between exclusively heterosexual and bisexual women with regard to sexual partner choice and attraction measures,13 this “between” placement did not hold for the health-related outcomes assessed in our study. Health measures for heterosexual women who had female sex partners did not fall consistently between those of bisexual and exclusively heterosexual women, nor did they consistently resemble any other group on substance use and sexual risk-taking and protection. This is consistent with findings on physical health concerns, in which heterosexual women with same-sex experience did not resemble any other orientation group, but indicated a higher number of physical health complaints than other groups.15
Whereas heterosexual women with a recent female sex partner had higher prevalences of substance use and some types of sexual risk taking, there is no evidence that their sex with other women directly accounted for these prevalences. Indeed, 69% of heterosexuals who had past-year female sex partners had only one such partner in their lifetimes. Those with less recent same-sex experience were also likely to have only one female sex partner. Rather, it appears that among heterosexually identified women, sex with another woman was likely to co-occur with higher numbers of male partners, higher sexually transmitted infection risk from male partners, and higher rates of substance use. More information is needed on the frequency and context of their sexual encounters with both men and women. The current analysis was unable to provide information on whether same-sex activity among women occurred while sober or high, one-on-one or in the context of nondyadic sex with male partners, or whether it occurred in the context of the sex trade. Moreover, data were not available to assess whether a female sex partner represented a single sexual encounter or an ongoing intimate relationship.
Though similar to heterosexual-identified MSM in terms of reporting a heterosexual identity while having engaged in sex with same-sex partners, caution is warranted in conflating these two groups as male and female sexualities differ biologically, psychologically and socially. Sexual minority identities develop earlier for men and are more stable over time. In contrast, women who are not exclusively attracted to either men or women often maintain stable attraction levels, but shift identities according to their relationship context or evaluation of future relationship potential.11,12 Men and women also respond differently to same-sex sexual advances,26 and sexual desire in women appears to be more fluid and context-dependent.12 Furthermore, some MSM have historically sought sex partners in venues such as bathhouses, parks, adult bookstores, restrooms, and the Internet.27 Such venues rarely exist for women, and when women do engage in sex-seeking behaviors, they are noticeably different in their approaches. Women are more likely to chat about sex on-line, whereas men are more likely to view pornography, and women are more likely to seek serious relationships.28
Because of sex differences in behaviors and in the development and maintenance of sexual minority identities, theories and models developed based on heterosexual MSM are not directly applicable to women. For example, concerns over nondisclosure of sexual behavior by MSM to female partners led to a proliferation of work based on “the down low” (DL), a concept developed to apply to heterosexual MSM, particularly men of color. This concept and has been critiqued as having impaired public health measures for black men by reinforcing the social construction of black sexuality as excessive, deviant, diseased, and predatory,29 and it would be imprudent to apply this or any concepts or research findings from MSM to heterosexual women with female partners. Whereas this study was unable to provide any information on whether heterosexual women who have had a female partner want to keep their same-sex activities or relationships secret, black women did not differ significant from white women in the sexual orientations they indicated, and do not appear less likely to develop a sexual minority identity.
Homosexual behavior in heterosexual women may be understood as indicative of a more sexually adventurous or experimental approach. Sensation seeking has been demonstrated to have power in explaining associations between substance use and HIV-related sexual risk taking in gay and bisexual men,30 but was not specifically assessed in the NSFG. Our findings of high prevalences of nonmonogamous male partners, high male partner numbers, and high rates of substance use are consistent with such an understanding, and may suggest situational same-sex sexual attraction and behavior in the context of greater sexual flexibility and substance use. Moreover, female homosexuality is less stigmatized and evokes fewer negative responses than male homosexuality, particularly among heterosexual men.31,32 Nevertheless, stigma continues to exist. Moreover, some of the “positive” responses to female homosexuality from males are in fact sexualized responses. It is plausible that for some women, conflict between their sexual behavior or attraction and their heterosexual identity could lead to substance use as a coping strategy. Further research is needed to understand sex seeking, sexual disclosure, stigma, the role of sensation seeking, and the contexts in which heterosexual women engage in same-sex sexual activities, or in which women who have sex with other women adopt or maintain heterosexual identities.
Whereas these data are cross-sectional and do not provide information on temporal relationships and changes, it is interesting to note that for all substance use measures and most sexual risk measures, heterosexual women with nonrecent female sex partners had prevalences that fell between those of heterosexual women with recent female partners and those with only male partners. Whereas statistically significant differences between heterosexual women with past-year versus less recent female sex partners for all substances other than marijuana disappeared in adjusted analysis, no changes for sexual health variables occurred with adjustment. This suggests that other than for marijuana use, differences between these 2 groups on substance use—but not on sexual risk—may be accounted for by differences in age or relationship status.
With the exception of these specific comparisons, adjustment had little effect on findings, and it thus appears that observed differences are not due primarily to demographic variation between groups. From this study, it appears that same-sex sexual behavior in heterosexual women—whether recent or remote—may serve as a marker for higher likelihood of recent and historical sexual risk with male partners and recent substance use. In clinical care and prevention settings this information may help identify women who should be screened for these issues, so that their education, prevention and health care needs can be appropriately met.
This research arose from earlier analyses on methods of measuring sexual orientation, and has further implications in this area. Heterosexual women may compose the largest sexual orientation identity group in population-based studies of women with female sex partners. Thus, population-based studies that classify sexual minority women exclusively on sexual behavior should not be used uncritically in making inference to lesbians, bisexuals, or other women with sexual minority identities, as the findings may be driven by the larger numbers of heterosexual-identified women. For example, while an association between same-sex sexual behavior and high numbers of male partners had previously been demonstrated using NSFG data,33 this was not homogenous among women with differing orientation identities. In our analysis, homosexual women had the lowest median number of lifetime male partners (med = 1) and were no more likely to have high numbers of male partners than exclusively heterosexual women. Bisexual women had higher numbers of male partners (med = 7), but heterosexually identified women with a recent or more remote female sex partner were significantly more likely than any other group to have high numbers of male partners. Future research is needed on the effects of grouping heterosexual women with sexual minority-identified women in health research. Additional work is needed to build understanding of the complex intersections of sexual behavior, identity, and attraction in women versus men, and to elucidate the context, frequency and implications of same-sex sexual experience in heterosexual women.
1. Laumann EO, Gagnon JH, Michael RT, et al. The Social Organization of Sexuality: Sexual Practices in the United States. Chicago: University of Chicago Press, 1994.
2. Ross MW, Essien EJ, Williams ML, et al. Concordance between sexual behavior and sexual identity in street outreach samples of four racial/ethnic groups. Sex Transm Dis 2003; 30:110–113.
3. Goldbaum G, Perdue T, Wolitski R, et al. Differences in risk behavior and sources of AIDS information among gay, bisexual, and straight-identified men who have sex with men. AIDS Behav 1998; 2:13–21.
4. Siegel K, Schrimshaw EW, Lekas HM, et al. Sexual behaviors of non-gay identified non-disclosing men who have sex with men and women. Arch Sex Behav 2008; 37:720–735.
5. Rietmeijer CA, Wolitski RJ, Fishbein M, et al. Sex hustling, injection drug use, and non-gay identification by men who have sex with men: Associations with high-risk sexual behaviors and condom use. Sex Transm Dis 1998; 25:353–360.
6. Pathela P, Hajat A, Schillinger J, et al. Discordance between sexual behavior and self-reported sexual identity: A population-based survey of New York City men. Ann Intern Med 2006; 145:416–425.
7. Harawa NT, Williams JK, Ramamurthi HC, et al. Sexual behavior, sexual identity, and substance abuse among low-income bisexual and non-gay-identifying African-American men who have sex with men. Arch Sex Behav 2008; 37:748–762.
8. Doll LS, Peterson LR, White CR, et al. Homosexually and nonhomosexually identified men who have sex with men: A behavioral comparison. J Sex Res 1992; 29:1–14.
9. Bauer GR, Welles SL. Beyond assumptions of negligible risk: Sexually transmitted diseases and women who have sex with women. Am J Public Health 2001; 91:1282–1286.
10. Marrazzo JM, Coffey P, Bingham A. Sexual practices, risk perception and knowledge of sexually transmitted disease risk among lesbian and bisexual women. Perspect Sex Reprod Health 2005; 37:6–12.
11. Diamond LM. Female bisexuality from adolescence to adulthood: Results from a 10-year longitudinal study. Dev Psychol 2008; 44:5–14.
12. Diamond LM. Sexual Fluidity: Understanding Women's Love and Desire. Cambridge, MA: Boston University Press, 2008.
13. Thompson EM, Morgan EM. “Mostly straight” young women: Variations in sexual behavior and identity development. Dev Psychol 2008; 44:15–21.
14. Koh AS, Gomez CA, Shade S, et al. Sexual risk factors among self-identified lesbians, bisexual women, and heterosexual women accessing primary care settings. Sex Transm Dis 2005; 32:563–569.
15. Cochran SD, Mays VM. Physical health complaints among lesbians, gay men and bisexual and homosexually experienced heterosexual individuals: Results from the California quality of life survey. Am J Public Health 2007; 97:2048–2055.
16. Kerker BD, Mostashari F, Thorpe L. Health care access and utilization among women who have sex with women: Sexual behavior and identity. J Urban Health 2006; 83:970–979.
17. Scheer S, Parks CA, McFarland W, et al. Self reported sexual identity, sexual behaviors and health risks: Examples from a population-based survey of young women. J Lesbian Studies 2003; 7:69–83.
18. Bauer GR, Jairam JA. Are lesbians really women who have sex with women (WSW)? Methodological concerns in measuring sexual orientation in health research. Women Health 2008; 48:383–408.
19. Trocki KF, Drabble L, Midanik L. Use of heavier drinking contexts among heterosexuals, homosexuals and bisexuals: Results from a National Household Probability Survey. J Stud Alcohol 2005; 66:105–110.
20. Drabble L, Midanik LT, Trocki KF. Reports of alcohol consumption and alcohol-related problems among homosexual, heterosexual and bisexual respondents: Results from the 2000 National Alcohol Survey. J Stud Alcohol 2005; 66:111–120.
21. Bailey JV, Farquhar C, Owen C, et al. Sexual behaviour of lesbians and bisexual women. Sex Transm Infect 2003; 79:147–150.
22. Friedman SR, Ompad DC, Maslow C, et al. HIV prevalence, risk behaviors, and high-risk sexual and injection networks among young women injectors who have sex with women. Am J Public Health 2003; 93:902–906.
23. Boehmer U, Case P. Physicians don't ask, sometimes patients tell: Disclosure of sexual orientation among women with breast carcinoma. Cancer 2004; 101:1882–1889.
24. Goodenow C, Szalacha LA, Robin LE, et al. Dimensions of sexual orientation and HIV-related risk among adolescent females: Evidence from a statewide survey. Am J Public Health 2008; 98:1051–1058.
25. Lepkowski JM, Mosher WD, Davis KE, et al. National Survey of Family Growth, Cycle 6: Sample design, weighting, imputation, and variance estimation. Vital Health Stat 2 2006; 142:1–82.
26. Beres MA, Herold E, Maitland SB. Sexual consent behaviors in same-sex relationships. Arch Sex Behav 2004; 33:475–486.
27. Grov C, Parsons JT, Bimbi DS. Sexual risk behaviors and venues for meeting partners: An intercept study of gay and bisexual men in LA and NYC. AIDS Behav 2007; 11:915–926.
28. Albright JM. Sex in America online: An exploration of sex, marital status, and sexual identity in internet sex seeking in its impacts. J Sex Res 2008; 45:175–186.
29. Ford CL, Whetten KD, Hall SA, et al. Black sexuality, social construction and research targeting ‘the down low’ (‘the DL’). Ann Epidemiol 2007; 17:209–216.
30. Kalichman SC, Heckman T, Kelly JA. Sensation seeking as an explanation for the association between substance use and HIV-related risky sexual behavior. Arch Sex Behav 1996; 25:141–154.
31. Herek GM, Capitanio JP. Sex differences in how heterosexuals think about lesbians and gay men: Evidence from survey context effects. J Sex Res 1999; 36:348–360.
32. Herek GM. Heterosexuals' attitudes toward bisexual men and women in the United States. J Sex Res 2002; 39:264–274.
33. Mosher WD, Chandra A, Jones J. Sexual behavior and selected health measures: Men and women 15–44 years of age, United States, 2002. Adv Data 2005; 362:1–55.