Despite the plethora of research that exists on the health of gay men since the AIDS epidemic, little research looks at the health of lesbians. The operationalization of lesbian health is no simple task. Conceptually it is difficult to define; when we refer to "lesbian health" we seem to set up a categorical difference between the health of lesbians and the health of heterosexual women. This raises the question for many researchers of whether we should even be studying lesbians separately from straight women where health is concerned. In this paper, I will raise some of these conceptual problems in researching lesbian health. Briefly, I will summarize some of the pertinent research on lesbian health as well as research on lesbians' experiences in the health care system. I will also discuss some of the methodological issues involved in lesbian health care research, especially the problems of categorizing which women fit into the definition of lesbian. Finally, I will detail some directions for future work on lesbian health.
Recent reviews of lesbian health care research show that many health care practitioners including physicians, nurses, and students, hold negative attitudes toward lesbians and gay men (see Stevens, 1992). Over two decades of research demonstrate consistent findings. For example, a 1986 study of 1009 medical doctors in San Diego showed that 40% of physicians were uncomfortable with treating lesbians and gay men, and 30% opposed admitting lesbians and gay men to medical school (Mathews, Booth, Turner, & Kessler, 1986).. Furthermore, in a 1989 study of 100 midwestern nurses teaching in a bachelor of science in nursing program, over half of nurses said that lesbianism was unnatural, over one-third said that lesbians were disgusting, and about one-fourth said that lesbians molest children (Randall, 1989). Furthermore, about thirty percent of these nurses said they would be uncomfortable providing care to lesbian patients. Despite their training, these nurses also indicated medically inaccurate beliefs about the health of lesbians. For example, 20% of nurses reported that lesbians were responsible for the transmission of AIDS in the general population. Similar attitudes among health care practitioners have been reported in two other samples. Among 120 nurses, 28% said that lesbians transmit AIDS (Eliason & Randall, 1991) and among 278 nursing students, 28% said that lesbians were a high risk group for AIDS, presumably because they were unable to distinguish lesbians from gay men (Eliason, Donelan, & Randall, 1992). Of these nursing students, about 13% also reported that lesbians were responsible for spread of other sexually transmitted diseases.
Given these types of widespread attitudes, it should come as no surprise that a high proportion of lesbians delay seeking health care because they fear the negative consequences of disclosing their sexual orientation. Across numerous studies, lesbians have reported hostility, demeaning jokes, anxiety, and excessive curiosity on behalf of their health care providers (see Stevens, 1992). On the more severe end of things, lesbians have reported being subjected to breached confidence, stereotyped comments, and even mental health referrals from their doctors. In addition, lesbians' nondisclosure of their sexual orientation has been shown to lead to inaccurate health teaching and misdiagnosis by their health care providers. Thus, lesbians tend to avoid the traditional health care process because they fear that the quality of their health care would be adversely affected if their sexual orientation became known. This point was exemplified in a recent study of 1000 lesbians in Los Angeles (Saunders, Tupac, & MacCulloch, 1988). Researchers found that women were more likely to seek help for a health problem from lesbian friends versus health care providers. Obviously, the avoidance of health care poses a threat to the physical well-being of lesbians. At least one study reports lesbians having fewer routine gynecological care visits and regular Pap tests compared to heterosexual women (Buenting, 1992). Such delays may lead to increased risk for delayed detection of cervical dysplasia (which is the abnormal growth of cells in that region and a condition that can lead to cervical cancer).
Although data are available regarding the health care experiences of lesbians, there is a paucity of information on the actual physical health among lesbians. One problem is that investigations into lesbian health are simply not funded at the governmental level. The little research that does exist on lesbian health has been conducted primarily by grassroots organizations either who pay for the studies out-of-pocket or who receive limited funding from supportive women's organizations. The data from these studies have been collected on mainly white, highly educated, young lesbians who typically were recruited to fill out health surveys at music festivals, gay pride parades, or at women's bookstores. Clearly, data gathered from this kind of research do not allow for generalizations across what we believe to be the lesbian population. We know almost nothing about the health of lesbians of color, rural lesbians, bisexual women, or older lesbians. Despite the obvious limitations, these type of data have been used by some researchers to make sweeping claims about lesbian health.
For example, an increasingly popular statistic is that the risk for lesbians to develop breast cancer is one in three. This contrasts sharply with the well-known statistic reported in the medical literature that one in eight women women (presumably straight), will develop breast cancer in their lifetime. This one in three statistic for lesbians was posited by Susanne Haynes of the National Institute of Health at the 1992 National Gay and Lesbian Health Conference and at the 1992 American Public Health Association Conference. However, Dr. Haynes did not base this number off of solid epidemiological data that she collected among lesbians. Instead, she examined at several studies, some of which included lesbians, some of which did not, and found that nullparity (not bearing children), was a risk factor in developing breast cancer. She suggested that since the majority of lesbians do not have children, lesbians thus had a greater risk of developing breast cancer compared to heterosexual women. However, no solid data are available on rates of child bearing among lesbians. Haynes also suggested that there were other "lesbian specific" behaviors that put them at greater risk for breast cancer which included: a higher rate of alcohol abuse and smoking, a higher rate of obesity, and being less likely to do routine breast exams and to seek routine gynecologic health care.
There are several problems with these assertions. Again, the first problem is that few studies actually compare the health behaviors of lesbians with the health behaviors of heterosexual women. Only then could one assert that lesbians actually have lower rates of nullparity than heterosexual women, or that they are more obese, or smoke more than heterosexual women. But before we can even begin to address this issue, researchers need to consider the question of forcing subjects to label themselves as "lesbian" or "straight".
Most lesbian health studies select only "out" lesbians to a large extent. For example, in one of the largest national studies of lesbians, women were asked to complete a questionnaire that had "The National Lesbian Health Survey" written across the cover. When researchers ask our subjects to label or to categorize themselves, they may completely miss the health problems of women who partner with women but who do not openly identify as lesbians. These women might be closeted and either do not attend lesbian events in the community (thus missing the opportunity to participate in research), or they do not label themselves as lesbian and would never fill out a lesbian health survey.
Another problem with Haynes' conclusions about lesbians is in how she identifies the variable that puts women at risk for breast cancer. She appears to be saying that lesbianism puts women at risk for breast cancer. However, closer inspection of her assertions reveal that it is not lesbianism that puts women at risk for breast cancer, rather the behavior of not bearing children. Haynes' research highlights the need to concentrate on the actual risky behaviors that people engage in when investigating health risk for any type of disease. For example, if we are worried about risk for gynecological disease, researchers might assume that lesbians would be at a lower risk than heterosexual women because they supposedly do not have sex with men. However, the researcher who only inquires about labels and forces subjects to categorize themselves without asking about sexual behaviors will completely miss the women who do label themselves lesbians but still have sex with men. Or, she will miss the women who label themselves as bisexual because they occasionally have desire for men, but actually only have sex with women.
Haynes makes the mistake of claiming that identity correlates or even leads to specific behaviors. Although no research has been published yet on the relationship between labeling of sexual orientation and actual sexual behavior, one might expect a moderate positive correlation--but not a one-to-one correlation as some researchers would have us believe.
At this point, it might be useful to briefly comment on some of the work in queer studies that has looked at the issue of identity and the problems of defining identity. Ed Cohen, in his essay "Who are 'we': gay identity as a political (e)motion", notes that there are vast problems in the categories of sexual identification. The slippages between categories are exemplified by bisexuals, "closet cases", and more recently, lesbians who sleep with men--all which may confound supposedly straightforward, scientific, quantitative research. The issue of identity politics as studied in the humanities is wide, varied, and clearly well beyond the scope of this paper. However, the work of Michel Foucault, Judith Butler, Teresa de Lauretis and many others all prove relevant to discussing problems in researching lesbian health. All have discussed the complexity and diversity of gay and lesbian sexualities. Judith Butler argues that sexual "identity categories tend to be instruments of regulatory regimes" (1991, p.13). As researchers, we need to be alert to the problems of the categorical imperative in our own work.
Concentrating on the relationships among the labeling of sexual orientation, actual sexual behavior, and risk for physical health problems may enrich the data that we collect. For example, past research suggests that "being out" leads to a greater propensity for negative life events, from experiencing homophobia in the health care system to experiencing homophobia at the work place. Greater stress has certainly been linked to greater risk for physical illness and disease. But what about the stress associated with being closeted ? Some researchers suggest that being closeted may lead to its own kind of stress--and as of now we do not know this kind of stress might affect rates of smoking, alcohol use, obesity, or health care visits for women who partner with women.
In sum, research on lesbian health has been greatly limited by methodology--from poorly selected samples of lesbians to poorly conceptualized questions with regards to identity and behavior. I do not mean to imply that we should not ask lesbians about their health. We simply do not have enough solid baseline data about the health status of lesbians to even know if and how lesbians might differ from heterosexual women. What I am suggesting is that we move away from stressing only labels and categories and begin to assess behaviors among all women--because after all, lesbians are women first, and how they label themselves may not be as relevant to their health as are their actual behaviors.