Current methodologies allow collection of information on sexual orientation with sufficient precision to discover important relationships. Further, such questions have been used successfully in a number of research areas with different populations. For example, a question on sexual identity was included in the Nurses' Health Study II, a large national cohort study, without apparent loss of participation. Identity was the focus in this study because it was believed that sexual identity and social relationships, rather than sexual behavior, were more likely to be determinants of breast cancer and other core concerns of the study.
The committee recommends that consideration be given to including questions about sexual identity, behavior, and attraction or desire in ongoing and future federal studies. These would include, for example, studies in which an association between sexual orientation and health can be hypothesized or in which discrimination based on sexual orientation may result in differential access to health care services. As appropriate, multiple dimensions of sexual orientation should be assessed whenever possible. Further, the rationale for including each question should be addressed in the study. These studies include, but are by no means limited to, the National Health and Nutrition Examination Survey, the National Household Survey on Drug Abuse, the National Survey of Family Growth, the American Community Survey, and the Youth Risk Behavior Survey. Pilot studies are recommended to test the feasibility of including these questions, with careful attention given to protecting confidentiality and assessing response bias and its impact on disclosure.
The committee recommends that researchers submitting proposals for federally funded research, whether unsolicited R01s, responses to Program Announcements, or responses to Requests for Proposals, routinely evaluate whether they should include sexual orientation questions in their protocols, just as they would other sociodemographic variables. The National Institutes of Health (NIH) review groups should be encouraged to consider whether or not sexual orientation should be assessed in proposed studies, and recommend inclusion of this data field when it would strengthen the value of the results.
Recommendation 4: Researchers studying lesbian health should consider the full range of racial, ethnic, and soc-