Lesbians' Sexual History With Men: Implications for Taking a Sexual History

Background  Health care providers may not solicit a comprehensive sexual history from lesbian patients because of provider assumptions that lesbians have not been sexually active with men. We performed this study to assess whether women who identify themselves as lesbians have a history of sexual activities with men that have implications for receipt of preventive health screening.

Objective  To convey the importance for health care providers to know their patients' sexual history when making appropriate recommendations for preventive health care.

Methods  A survey was printed in a national news magazine aimed at homosexual men, lesbians, and bisexual men and women. The sample included 6935 self-identified lesbians from all 50 US states. The outcomes we measured were respondents' number of lifetime male sexual partners and partners during the past year, their lifetime history of specific sexual activities (eg, vaginal intercourse, anal intercourse), their lifetime condom use, and their lifetime history of sexually transmitted diseases.

Results  Of respondents, 77.3% had 1 or more lifetime male sexual partners, 70.5% had a lifetime history of vaginal intercourse, 17.2% had a lifetime history of anal intercourse, and 17.2% had a lifetime history of a sexually transmitted disease. Exactly 5.7% reported having had a male sexual partner during the past year.

Conclusion  These findings reinforce the need for providers to know their patients' sexual history regardless of their reported sexual orientation, especially with regard to recommendations for Papanicolaou smears and screening for sexually transmitted diseases.

ALTHOUGH THERE is strong consensus among primary care provider organizations that clinicians should obtain a sexual history from their patients, this aspect of a patient's medical history may not receive adequate attention because of reluctance on the part of the health care provider (ie, physician, nurse practitioner, physician assistant) or the patient.^1-8 Provider knowledge of their patients' sexual history is important for (1) determining when to screen for sexually transmitted diseases (STDs) and pregnancy, (2) recommending appropriate preventive health care services, (3) developing a differential diagnosis, (4) providing relevant risk-reduction counseling, and (5) determining whether a patient has medically treatable sexual problems that could be inhibiting a fulfilling sexual life.

Women who have a history of vaginal intercourse are at greater risk for gynecologic infection and malignant neoplasms than women who have never had vaginal intercourse, and this risk increases for women who have a history of vaginal intercourse without a condom.^9,10 Unless health care providers specifically ask a patient about her sexual history and the activities she has engaged in, they may make incorrect assumptions regarding her sexual history and sexual orientation, and these assumptions may affect the quality of health care services provided.

A lesbian is, by definition, a woman who is primarily physically and emotionally attracted to other women^11-15; however, the woman's past or current sexual partners may not be limited to women. Because of ambiguity of public perceptions and the fluidity of sexual activity, clinicians may assume, with good intentions, that they do not need to ask patients they know or assume to be lesbians about a history of male sexual partners. Results of previous studies indicate that many lesbians have had at least 1 male sexual partner at some time during their lives.^16-26 However, information has not been available on pertinent aspects of lesbians' sexual history with men, including age at onset of sexual activity with men, specific sexual activities, recency of sexual activity, number of lifetime male partners, and use of condoms.

In this article, we use national data from a magazine-based survey to determine whether women who identify themselves as lesbians have a history of engaging in vaginal intercourse and other sexual activities with men and, if so, the nature and duration of their sexual history with men, as well as their history of STDs and abnormal Papanicolaou (Pap) smears.

A 186-item questionnaire was developed by 3 health services researchers (including M.A.S. and J.L.) and was printed in The Advocate, a monthly national news magazine for homosexual men, lesbians, and bisexual men and women. The questionnaire appeared as a several-page insert in the center of the March 1995 issue and included a postage-paid, return-addressed mailer. The questionnaire included demographics, sexual orientation, sexual history with men, and gynecologic history, as well as other items not covered in this article.

Marketing and subscription information at the time of the survey indicated that 88 000 copies of each issue of The Advocate were distributed and that the female readership was about 24 000. There were 7929 respondents. The analysis sample for this article consisted of 6935 women from the 50 United States who identified themselves as lesbians on the basis of their response to the first item of the survey, "How do you describe your sexuality," with the following response options: (1) homosexual, gay, or lesbian; (2) bisexual; (3) heterosexual or straight; and (4) not sure. We only included lesbians because we were interested in assessing sexual activity with men among a population of women whose health care providers might make inaccurate assumptions about their sexual history on the basis of their orientation. Women who listed their orientation as bisexual or unsure (n = 862) and 132 women from 18 foreign countries were omitted from this analysis.

The study protocol was submitted to the university Human Subjects Protection Committee for the University of California, Los Angeles, and was granted institutional review board approval.

Respondents reported their lifetime sexual activities with men, including kissing, mutual masturbation, fellatio, cunnilingus, vaginal intercourse with and without a condom, anal intercourse with and without a condom, and number of male sexual partners ever and during the past year. In the survey, vaginal intercourse was defined as occurring with men. The survey also included age at first and most recent vaginal intercourse and history of gonorrhea, Chlamydia infection, trichomoniasis, pelvic inflammatory disease, syphilis, genital or anal herpes, genital or anal warts, and human immunodeficiency virus (HIV) infection. We classified respondents as having a history of at least 1 STD if they reported 1 or more of these conditions. The survey also included items regarding history of an abnormal Pap smear and history of HIV testing.

We performed bivariate analyses by means of χ² tests. We performed logistic regression for the following outcome variables: a history of vaginal intercourse, having a male sexual partner within the past year, and vaginal intercourse without a condom. We included in the regression models all independent variables that were significant at P<.05 in bivariate analyses.

In response to our survey, 84.8% of lesbian respondents were between 25 and 49 years of age (mean, 35 years), and 87.5% of respondents described themselves as white (Table 1). College graduates comprised 22.2% of the respondents, and an additional 40.6% had some graduate or professional school experience. An annual income between $20 001 and $50 000 was reported by 54.5%, and 35.3% lived in cities with greater than 1 million people.

Of the respondents, 77.3% reported ever having had a male sexual partner, 70.5% had engaged in vaginal intercourse at least once, and 17.2% had engaged in anal intercourse at least once (Table 2). Among women who had engaged in vaginal intercourse, the mean age at first intercourse was 18 years, and the mean age at last intercourse was 25 years. Our data showed that 5.7% of respondents reported having had 1 or more male sexual contacts within the preceding year. Respondents in this study also reported a history of ever having participated in the following sexual activities with men: kissing (94.5%), mutual masturbation (64.0%), fellatio (62.0%), and cunnilingus (62.3%).

Results of the multivariate logistic regression model are found in Table 3. Lesbians were significantly more likely to have a history of vaginal intercourse if they had not graduated from college and were less likely if they were younger than 50 years, or if they lived in small or medium-sized cities or rural areas.

The multivariate regression model that we used to assess the independent effects of the explanatory variables on whether lesbians had had a male sexual partner during the preceding year showed that women who were not white, who were younger than 50 years, who had not graduated from college, and who had an annual income of $20 000 or less (vs those with an annual income greater than $50 000) were more likely to have had 1 or more male sexual partners during the past year (Table 3).

Our results showed that 63.9% of all respondents and 88.2% of lesbians who had had vaginal intercourse reported that they had participated in vaginal intercourse without a condom, and these findings varied significantly in the bivariate analyses by race or ethnicity, age, education, and income (Table 2); also, 69.1% reported that they had ever used a condom during vaginal intercourse. For all respondents, 15.8% reported a lifetime history of anal intercourse without a condom, and there were significant differences by age, income, education, and place of residence. Our data showed that 4.8% reported ever using a condom during anal intercourse, with significant variation by race or ethnicity and income.

In the regression model, lesbians were significantly more likely to have engaged in vaginal intercourse without a condom if they were older than 50 years (compared with women younger than 25 years and women 25 to 29 years old), or if they reported more than 1 lifetime male sex partner (Table 3).

In response to our survey, 17.2% of respondents reported a lifetime diagnosis of an STD, and 17.3% reported a lifetime history of an abnormal Pap smear (Table 4). The most commonly reported STDs were trichomoniasis (6.0%), genital or anal warts (4.8%), Chlamydia infection (4.6%), genital or anal herpes (3.3%), pelvic inflammatory disease (2.0%), and gonorrhea (1.62%), with fewer than 1% of respondents reporting a history of syphilis (0.3%) or HIV infection (0.1%). Lesbians who reported 6 or more male sexual partners were most likely to have had an STD at some point during their lives. Lesbians who had participated in vaginal intercourse were more likely to have reported a history of an STD than lesbians who did not have a history of vaginal intercourse (21.4% vs 7.3%; P<.001). Lesbians who had engaged in vaginal intercourse and those who had engaged in anal intercourse reported higher rates of STDs and abnormal Pap smears than those who had not engaged in these activities (Table 4).

Of all respondents, 53.2% had had an HIV test at some point during their lifetime, and 66.1% of individuals who had a lifetime history of an STD had undergone testing for HIV (P<.001). Among women who had engaged in vaginal intercourse or anal intercourse without a condom, 42.3% and 37.1%, respectively, had never been tested for HIV. Overall, 5.9% of respondents had ever participated in anal intercourse without a condom and had never been tested for HIV, and 26.9% of respondents had ever engaged in vaginal intercourse without a condom and had never been tested for HIV.

In this study, a majority of lesbians had engaged in penile vaginal intercourse, with a significant proportion participating in vaginal intercourse without a condom. Younger women were more likely to have had a male sexual partner during the preceding year. By contrast, older women were more likely than younger women to have engaged in vaginal intercourse, perhaps because the former have had a longer time during which to engage in such activity. Older women were also less likely to have used a condom, which may indicate that public health messages about safer sex are not influencing them or that such messages were not yet available when they were having intercourse.

If clinicians assume that a woman who identifies herself as a lesbian has not had any sexual contact with men, or that such contact was only in the distant past, they may not make appropriate diagnostic and treatment recommendations. Of particular importance with regard to recommendations by providers for cervical cancer screening is lesbians' prior sexual history with men, including age at first vaginal intercourse, use of condoms, and number of lifetime male sexual partners. The clinical importance of these findings relates to clinicians' advice to their lesbian patients regarding receipt of appropriate health care services, including cervical cancer screening, testing for STDs, advice regarding safe sexual behaviors, and information about contraception and fertility.

The findings from our study are consistent with previous studies that show a history of sexual contact with men for many lesbians^16-26; however, our study also provides information on specific sexual activities that lesbians have engaged in with men and on their use of condoms. This information is important for clinicians in the provision of appropriate health care services to lesbians. In a more recent study, Lemp et al¹⁶ reported on the prevalence of risk behaviors for HIV, such as "unsafe sex with men," among a sample of 498 lesbians and bisexual women in San Francisco and Berkeley, Calif. Forty percent of these lesbians had engaged in unprotected vaginal (39%) or anal (11%) sex within the preceding 3 years, some with homosexual or bisexual men or male injecting drug users.¹⁶

According to guidelines from the US Preventive Services Task Force, regular Pap smears are recommended for all women who are currently, or have been previously, sexually active with men and who have a cervix, beginning at age 18 years or when the woman first engages in sexual intercourse.¹ After at least 3 consecutive annual tests have been normal, Pap smears may be done every 3 years for women at low risk for cervical dysplasia. These recommendations vary among specialty organizations.^27,28 In general, women who have engaged in sexual intercourse with 5 or fewer lifetime male partners and who have not participated in unprotected vaginal intercourse are characterized as being at lower risk for cervical dysplasia.¹ Women who become sexually active with a new male partner within a year of their last Pap smear are advised to undergo annual screening.

The Centers for Disease Control and Prevention established that unprotected anal intercourse is the highest-risk sexual activity for the transmission of HIV, and unprotected vaginal intercourse is also a high-risk sexual activity.^29,30 In our study, 5.9% of respondents had ever participated in anal intercourse without a condom and had never been tested for HIV, and 26.9% of respondents had ever engaged in vaginal intercourse without a condom and had never been tested for HIV.

Other clinical implications of a history of vaginal intercourse with STDs include decreased fertility, chronic infection (eg, herpes simplex virus), and risk of transmission of disease to sexual partners of either sex. If women are or were sexually active with men, the potential for infection may be greater than if they have only been sexually active with women³¹; however, there is evidence that infection with herpes simplex virus, Trichomonas vaginalis, Gardnerella vaginalis, and perhaps human papillomavirus, may be transmitted between female sexual partners.^26,32,33 The transmission of HIV between female sexual partners is less clear. There are several published case reports of women with HIV whose only known potential route of transmission was female-to-female sexual contact.^34-40 However, one short-term prospective study found no female-to-female transmission after 6 months among 18 couples discordant for HIV infection who participated in orogenital sex.⁴¹

This study has a number of important strengths. First, in contrast to previous studies, this study is based on a much larger number of respondents who resided in all 50 states, allowing subgroup analyses. Second, the questionnaire included detailed questions about types of sexual behaviors, which, combined with the large sample size and demographic information, permits analysis of relationships that have not previously been examined. Third, the sample includes a large number of women whose sexual behavior puts them in important health risk categories. Fourth, a broad-based media solicitation yields broader geographic diversity of representation that surpasses the convenience samples of lesbians studied to date.

Although this article presents results from the largest sample of lesbians studied to date, there are some limitations. This study did not use probability sampling, so generalization to all lesbians must be done with caution. The difficulty in identifying a probability sample partly results from the hesitancy of some women to disclose their sexual orientation because of fear of negative reactions from employers, family, and others.^21,42-44 Those women who read The Advocate magazine may not be representative of lesbians in general, and the lesbians who responded may not be representative of the full female readership of the magazine. How these sampling biases relate to either sexual history or health-seeking behavior is unknown in this population. Data suggest that individuals who answer surveys about sexual issues tend to hold more liberal sexual attitudes and be more sexually active than nonrespondents.⁴⁵ Magazine readership also tends to be skewed toward people with higher-than-average education and income, which was also true of our sample. Another limitation is the reliance on self-report for disease conditions, a limitation common to survey-based studies.

If health care providers are to make appropriate recommendations to their lesbian patients regarding preventive health care screening, preventive behaviors, and treatment options, they must obtain an accurate social history that includes a woman's sexual history as well as, but not limited to, any history of substance use, physical or emotional abuse, social support systems, and employment. In previous studies, lesbians have reported negative experiences with health care providers that included insensitive comments, inadequate or inappropriate health care, and refusal to provide treatment.^{18,20,21,23,24,46} When introducing the topic of the sexual history to the patient, the clinician can preface the questions with an explanation of why an accurate and complete sexual history is important, and then ask the questions without specifying a presumed sex of the woman's partners. For example, the clinician can explain that women who have had vaginal intercourse have an increased risk for cervical dysplasia, that cervical cancer screening is recommended even in women who are no longer sexually active with men, and that STDs may have long-term effects on fertility.

In summary, many lesbians have a history of sexual contact with men that includes unprotected vaginal intercourse, and some have engaged in unprotected anal intercourse. Therefore, clinicians should not assume that women who describe themselves as lesbians have never engaged in sexual activity with men or are not currently doing so. It is important for the clinician to know a patient's complete medical and social history, including current and past sexual activity, to make appropriate decisions regarding the provision of appropriate health care. This care includes, but is not limited to, performing Pap smears, screening for STDs, assessing HIV risk factors, and advising on sexual risk reduction.

Accepted for publication March 15, 1999.

This study was supported by grants from the Centers for Disease Control and Prevention, Atlanta, Ga, and the Lesbian Health Fund, San Francisco, Calif, and by the UCLA Robert Wood Johnson Clinical Scholars Program, Los Angeles, Calif, and the University of California, Los Angeles, National Research Service Award Primary Care Fellowship.

This work does not necessarily represent the opinions of the funding organizations or of the institutions with which the authors are affiliated.

We thank Carol Edwards for programming and Katherine Kahn, MD, and Lillian Gelberg, MD, MSPH, for their comments on drafts of the manuscript.

Reprints: Allison L. Diamant, MD, MSHS, Division of General Internal Medicine and Health Services Research, Department of Medicine, UCLA, 911 Broxton Plaza, Los Angeles, CA 90095 (e-mail: adiamant@ucla.edu).