Though medical care in the United States has promoted health and longevity, many disparities persist even after socioeconomic status and other factors have been accounted for 12. Sexual minorities are a group for whom disparities in health are prevalent and problematic 34. Lesbian, Gay, Bisexual, and Transgender (LGBT) individuals are often “invisible” due to the ability of members to hide their status and thus avoid bias because they may be construed as “different” from their peers 56. As a result of familial, societal and religious pressures, many LGBT individuals hide their status and relatively few make their orientation known to their health care providers. Those who disclose their orientation may find that their physician is unprepared and even unwilling to discuss same-sex relationships and behavior 7.

Medical educators recognize the need for physicians to understand and interact with patients of different cultural backgrounds since prejudices of medical students and physicians have demonstrable and important effects on patient wellness 8. Often, health care providers are not aware of their biases, which can remain unconscious. This leads to deleterious patient outcomes due to assumptions that one is heterosexual and negative caregiver attitudes that are not openly discussed in medical education settings 9. Social desirability bias often limits the disclosure of negative attitudes 1011. Health care providers who have negative attitudes toward same-sex behavior have been found to provide inadequate care for LGBT individuals 12. Furthermore, studies assessing prejudicial reactions of health care providers have been limited in recent years. In 1982, a questionnaire sent to members of the San Diego County Medical Society revealed that 23% of respondents exhibited prejudiced attitudes and 30% would reject a highly qualified gay applicant to medical school 13. Another study revealed “low-grade homophobia” among medical students that did not change significantly despite panel discussions and clinical experiences 14. A more recent look at San Diego County Medical Society members’ view on gay students revealed a much decreased prevalence of sexual prejudice, but one that still existed and was associated with phobias about human immunodeficiency virus infection 15.

The objectives of this study were to determine medical students’ attitudes toward gay male behavior, persons, civil rights, and male toughness. We chose to focus on gay men in the context of “masculinity” or “normative” views on male gender roles because there are no studies to date suggesting that these concepts should be discussed in medical school lectures. We hypothesized that individuals would not exhibit prejudice toward a person’s identity, but would exhibit bias when reflecting on the intimate practices of others. We sought to investigate the prevalence and correlates of negative attitudes toward gay identity and behavior. We predicted negative attitudes toward gay men among older students and males based on public surveys suggesting these populations have more aversive reactions 16.

251 out of 371 medical students responded (response rate: 68%). The histograms revealed a highly skewed, non-normal distribution, thus requiring non-parametric methods to determine significance. The results of the multiple imputations suggested that there were no large differences between any of the analyses, with p-values differing only in the third decimal place. For this reason, the first imputation was used and all results are from this complete data set.

Respondents had an average age of 27 years (range from 21 to 45 years): 94.8% of respondents were heterosexual. Of those who completed the survey, 59.4% of the respondents were women and 50.6% were Caucasian. Additionally, we delineated three age groups to determine whether age, irrespective of year of enrollment, influenced attitudes toward homosexuality. We did not ask respondents about their country of origin nor specific year in the program (Table 1).

^a Lesbian, Gay, Bisexual, or Transgender (LGBT).

Cronbach’s alpha coefficient showed that survey elements were moderately internally consistent across subscales of gay behavior (.86) and male toughness (.82). However, scale score reliability for persons and civil rights (.53 and .57, respectively) was less robust, perhaps due to the low number of questions in those subsections (Table 2). Nearly all respondents endorsed positive attitudes toward gay persons and their civil rights, though skewed data was obtained for the behavior and male toughness scales.

^*Reverse scored for scale calculation (values of 1, 2, or 3 corresponded to 7, 6, or 5 respectively).

However, for some respondents, the thought of two men holding hands or having sex was more “disgusting” than the thought of a man and a woman engaging in the same acts; a significant minority of students found homosexuality to be an unnatural form of sexual expression. A significant portion of respondents exhibited traditional views on male gender roles with respect to concepts of toughness and aggression (Table 2).

We next sought to determine if respondents from different demographic groups scored differently. The results evidence disparities in response according to respondents’ demographic characteristics (Table 3).

This study provides a comprehensive characterization of attitudes toward gay men endorsed by students at a large U.S. medical school comprising a diverse student body. The majority of respondents were affirming of gay men and same-sex behavior. Overt disgust towards gay men was infrequent. However, most questions in the “behavior” scale had wide standard deviations suggestive of diverse attitudes. Substantial minorities of students expressed disgust to gay male behavior that correlated moderately with negative attitudes regarding civil rights and normative notions of male toughness. The results of this study reveal the need for assessing and revising current methods of or approaches to instruction in the care of sexual minorities.

Good doctor-patient relationships require trust and mutual respect. Clinical heterosexism, the assumption that the patient is heterosexual, interferes with the formation and maintenance of a healthy doctor-patient relationship, and important opportunities to engage a gay patient in healthful behaviors are likely to be missed 27. Though typically covert, biases such as those reported in this study can lead to subconscious actions. Prejudices have been shown to result in overt antigay behaviors 28. More worryingly, disgust toward gay men experienced by health care providers and respondents to our survey is particularly powerful 29. Such visceral responses are likely to undermine quality of care and reinforce stigma.

Homosexuality has been considered to be “a natural difference, like left-handedness.” 30 However, when specifically asked, nearly one-third of medical students responded either negatively or ambivalently to the statement that male homosexuality is as natural a form of expression as heterosexuality. Our study demonstrates that some medical students find gay men and their behavior confusing. Studies have demonstrated the direct relationship between implicit and explicit measures of bias toward gay men that is most apparent among heterosexual men and based upon affective responses. Our study measured explicit bias. By its nature, unconscious bias goes unrecognized by people who see themselves as tolerant or at least hoping to “extract conformity with social norms.” 31 Though individuals may feel “comfortable” with LGBT issues in theory, when they are explicitly presented them, it becomes disconcerting. Educational modules must therefore explicitly endorse the “normality” of homosexuality.

Our study has several strengths. The survey had a high response rate. Despite this, it is possible that those who did not respond are less comfortable with discussion of sexually and prejudice. Our results were consistent with a related study on exposure to LGBT patients conducted among medical students at the New York University School of Medicine 3233. Our study sought additional components of beliefs about gay men to more comprehensively evaluate for the presence of bias. Since California is historically politically liberal, it is likely that a similar survey administered in a more conservative part of the country would find a higher prevalence of prejudicial beliefs. Although the University of California, Davis, School of Medicine, is a committed leader in the inclusion of LGBT issues in the curriculum, a lack of acceptance was found among a significant minority of students, thus highlighting the need to help students recognize and understand their own biases for better communicating in a socially- and culturally-appropriate manner with patients from all sexual orientations, races, ethnicities, and so forth. There was a discrepancy between the percent of students that let the school administration know they identified as LGBT (2.3%) and those who identified as such on the survey (5.3%). This underscores the notion that some students are more comfortable with their sexuality while others feel less forthcoming.

Several limitations should be considered when interpreting our findings. The results represent the views of a single medical school student population within the United States to a single stigmatized group. Our study did not evaluate a respondent’s year in medical training to determine whether attitudes changed with increased amount of time in school. Most medical students are in their twenties, making conclusive statements regarding age and negative attitudes more difficult because the range is more narrow. The cross-sectional and correlational analyses preclude a broader generalization of how medical students in the United States and globally frame negative responses to gay men, and does not characterize their perceptions of lesbians, bisexual or transgender people, or intersex persons. Given that many questions used the word “disgust,” it is possible that even higher rates of bias would be seen if less extreme words such as “discomfort” or “unease” were used instead. In this study, we utilized explicit measures of bias. The questions exposed self-reported bias, which may not correspond to privately held beliefs. Some individuals may exhibit negative reactions to those different from themselves, but will not divulge this information when directly asked 34. Future work can augment this study to assess the relationship to beliefs that are explored through measurement of implicit measures to determine if there are discrepancies among students in higher education, where disguise of bias may be more apparent 35.

Efforts to address prejudices, among other biases that may be covert, are needed to facilitate open discussions and thereby improve care. Previous research has shown that individuals who believe that homosexuality is “congenital” exhibit more positive attitudes toward sexual minorities. Labeling or constructing groups as different in speech, appearance, or socio-cultural background has deleterious effects on the ability to provide sound and culturally-adept treatment 363738. Several innovative workshops have been developed that help to address the rich needs of this underserved group beyond sexual practice and pathology but more are needed 39. This includes tackling concepts of gender roles and confronting patient issues that would otherwise be unspoken.

We show that negative attitudes toward gay behavior were linked to more normative views on male gender roles, or how men are expected to act. Our study advances medical education for gay men by supporting the need for lectures to discuss masculinity due to documented correlations with heterosexism 40. Changing medical student attitudes cannot be done in isolation and should be a component of a comprehensive organizational approach that involves leadership and faculty role models to shape the culture throughout all the clinical and academic venues encountered by students.

There is considerable evidence that competency in dealing with populations experiencing health disparities require systematic development of a competency-based curriculum throughout all of medical school. Developing a comprehensive four-year curriculum around LGBT health is a critical first step. However, even the presence of LGBT patients and lectures disseminating facts may not adequately address underlying biases. More interactive methods that incorporate diverse fields of study, including anthropology and literature, could be used to invoke empathy. Incorporation of interdisciplinary efforts in current educational modules may attenuate biases seen not only towards sexual minorities, but the underserved from various ethnic and social backgrounds.

Our study demonstrates that sexual prejudice is more likely to be found in male and younger medical students. Because the overall data is not normative, it is difficult to determine whether there was an overall “positive” or “negative” prejudice. The first finding is surprising given that most polls on homosexuality show a more positive attitude among younger respondents. Students may acquire more positive attitudes toward sexual minorities as exposure to LGBT patients and other life experiences increases through medical school. Though attitudes toward gay behavior are complex, equality in the provision of clinical care mandates improved educational interventions. Health professionals must acknowledge biases they have despite discomfort this may cause. The survey utilized in this study could be used at other medical schools to assess the extent of sexual prejudice against gay men and therefore provide impetus to implement curricular changes in concert with a cultural competence committee. Longitudinal studies with specific interventions should be explored to determine whether prejudice reduction strategies could improve prejudicial reactions using introspection and facilitated clinical encounters. Efforts to increase opportunities to confront and reflect on biases are needed to avoid perpetuating a prevalent and problematic stigma.

(LGBT), Lesbian, Gay, Bisexual, Transgender.

The authors have no competing interests, financial or otherwise.

KM designed the study, participated in dissemination and data collection, and drafted the manuscript. RLK and GTM interpreted the data and drafted the manuscript. MDW carried out the statistical analyses and drafted the manuscript. FTF oversaw the study and drafted the manuscript. All authors read and approved the final manuscript.