Sexual orientation and associated characteristics among North American academic psychiatrists.

Over the past several decades, a number of studies have investigated the psychological and sociological characteristics of mental health practitioners (Armor & Klerman, 1968; Bodkin, Klitzman, & Pope, 1995; Monk & Thomas, 1973). These studies have found wide variations among mental health practitioners in attitudes and approaches toward psychiatric treatment and etiology. Over time, these findings have documented a shift among professionals from psychological approaches, dominated by Freudian perspectives, to biological approaches and theoretical understandings of disorders. None of these studies, however, have assessed whether differences in sexual orientation among mental health professionals are associated with differences in treatment approaches, attitudes, and perspectives, or other professional and personal characteristics.

This issue is of particular interest because research over recent years on the etiology of homosexuality has shifted from psychological to biological explanations and has received much attention in the popular press and elsewhere (Gladue, Green, & Hellman, 1984; Hamer, Hu, Magnuson, Hu, & Pattatucci, 1993; Levay, 1991; Pillard & Bailey, 1995). Controversy on this issue continues among gays and lesbians, with many preferring social constructionist explanations over biological explanations of sexual orientation.

In light of these issues, it is not clear whether gay psychiatrists, given the stigma they have faced and their perceptions of not fitting in with "straight" society, view or treat psychiatric illness differently (e.g., more or less biologically) than their colleagues. It is also not clear whether gay psychiatrists are more or less concerned with orthodoxy, socially desirable behaviors, or humanistic approaches than are their heterosexual colleagues. These considerations are of interest, as gays and lesbians have become more visible as a group among mental-health providers and patients. Clinics specializing in the mental-health needs of the gay and lesbian community have been established in cities across the country. Increasing numbers of mental-health practitioners have also identified themselves as gay and lesbian and have written about their perspectives in the field (Isay, 1989, 1995). Yet, to date there have been no studies comparing gay and lesbian psychiatrists to their straight colleagues in professional or other personal characteristics.

We examined these issues among academic psychiatrists because they are at the forefront of biological and other recent advances in the field. Moreover, academic psychiatrists are an important group, as they very much influence the direction of the field by teaching medical students and residents, conducting research, and editing journals. We chose academic institutions located in different regions and in different types of geographic areas to sample broadly within this population.

METHODS

We sent a brief, anonymous questionnaire to 972 North American academic psychiatrists at five leading academic medical centers in 1990. The survey instrument (see Appendix) contained questions concerning demographic features, various personal and professional attributes, and assessments of biological versus psychological orientation. To minimize response bias, we did not inform respondents of our specific hypotheses. We stipulated in a cover letter that to protect confidentiality and anonymity, we would not conduct any analyses by individual institution.

In the first publication to report findings of this study, we compared attributes of biologically oriented and psychologically oriented psychiatrists (Bodkin et al., 1995). We found that almost two thirds of respondents could be classified as either biological (27%) or psychotherapeutic (37%) in orientation, according to the proportion of their caseload to which they provided psychotherapy ([is less than or equal to] 25% vs. [is greater than] 75%). Compared with biologically oriented respondents, psychotherapeutically oriented respondents more frequently reported personal histories of psychiatric disorders (64% vs. 39%) and greater satisfaction with clinical work (81% vs. 53% reported that they were very satisfied). Biologically oriented respondents were also more likely to be younger. Subsequent comparisons corrected for age found that biologically oriented psychiatrists were more likely to engage in research as a primary or secondary activity, and to have pursued research since completing residency training. Psychotherapists as a group were more likely to have experimented with marijuana, to be women, to be divorced, and to have family histories of psychiatric illness in one or more first-degree relatives.

One of the questions on the survey instrument regarded sexual orientation. The respondent could self-identify as exclusively heterosexual, predominantly heterosexual, bisexual, predominantly homosexual, or exclusively homosexual. For the present report, we compared the exclusively heterosexual group with the pooled remaining groups on all of the other variables assessed in the survey. The significance of the comparisons between the two groups was assessed by Fisher's exact test (two-tailed).

RESULTS

Of the 972 psychiatrists surveyed, 435 completed and returned the questionnaire, and 76 questionnaires were returned as undeliverable, yielding a response rate of 49%. Of the respondents, seven (1.6%) did not answer the question regarding sexual orientation. Of the remaining 428 respondents, 389 (90.9%) reported being exclusively heterosexual, 15 (3.5%) predominantly heterosexual, 6 (1.4%) bisexual, 7 (1.6%) predominantly homosexual, and 11 (2.6%) exclusively homosexual.

Significant associations between sexual orientation and other survey variables are summarized on Table 1. Exclusively heterosexual psychiatrists were more likely than nonexclusively heterosexual psychiatrists to be Jewish, to have reported more first-degree relatives with a DSM-III-R Axis I psychiatric diagnosis (American Psychiatric Association, 1987), to prescribe medications to at least some of their patients, and to have engaged in research after residency. However, exclusively heterosexual psychiatrists were significantly less likely to report personal use of illicit drugs as a whole and hallucinogens in particular. In addition, there were trends for the exclusively heterosexual psychiatrists to have used less marijuana (any marijuana use: 58% of exclusive heterosexuals vs. 74% of others, p = .059) and stimulants (any stimulant use: 18% of exclusive heterosexuals vs. 32% of others, p = .052).

Table 1. Sexual Orientation and Characteristics of North American Academic Psychiatrists

                                         Sexual Orientation

                                    Exclusively   Not Exclusively
Characteristics                     Heterosexual   Heterosexual

Marital Status (Single)                 5%              49%
                                       (18)            (19)
Non-Jewish                              54%             79%
                                       (207)           (31)
Protestant                              30%             54%
                                       (114)           (21)
Number of First-Degree Relatives
 With Psychiatric Diagnosis:
                0                       64%             72%
                                       (248)           (28)
                1                       25%             18%
                                       (98)             (7)
                2                       8%              0%
                                       (33)             (0)
                3                       2%              10%
                                        (6)             (4)
                4                       1%              0%
                                        (3)             (0)
                5                       .3%             0%
                                        (1)             (0)

Prescribing No Medications              3%              13%
                                                        (5)

Any Drug Use                            58%             77%
                                       (224)           (30)

Any Hallucinogen Use                    17%             32%
                                       (63)            (12)

Research Post-Residency                 65%             47%
                                       (252)           (18)

Characteristics                       p value

Marital Status (Single)                .000

Non-Jewish                             .002

Protestant                             .003

Number of First-Degree Relatives       .015
 With Psychiatric Diagnosis:
                0

                1

                2

                3

                4

                5

Prescribing No Medications             .017

Any Drug Use                           .025

Any Hallucinogen Use                   .027

Research Post-Residency                .034

Note. All comparisons are performed by Fisher's exact test (for 2 x 2 comparisons) or by exact test for 2 x 6 comparisons (for number of first-degree relations with a psychiatric diagnosis).

No trends or statistically significant differences were found between the two groups in age (as measured in years or in ten-year intervals), gender, college major, plans to enter psychiatry when beginning medical school, consideration of other medical specialties, satisfaction with medical school, principal activities in psychiatry, percentage of patients provided formal psychotherapy, political views, satisfaction with clinical practice, or consulting a mental-health professional for a specific diagnosis. Overall orientation (biological vs. psychological), as assessed by several of the questions, was almost identical among exclusively heterosexual and not exclusively heterosexual psychiatrists. A significant difference emerged on only one item for biological orientation: The exclusively heterosexual psychiatrists were less likely to report a clinical practice in which they prescribed medications to none of their patients (see Table 1).

In a further analysis, we found that 90% of the psychiatrists who were not exclusively heterosexual were male, whereas 79% of the exclusively heterosexual psychiatrists were male. Given the preponderance of males, we duplicated the previous analyses restricted to males, assessing differences between exclusive and non-exclusive heterosexuals, The results of this analysis were almost identical to that of primary analysis, with virtually all the previously significant differences retaining a significance level of p [is less than] .05. The only exception was for any hallucinogen use, which lost some statistical power (p = .067 among males only). In addition, one variable became significant: Of the males who were not exclusively heterosexual, 69% had consulted a mental-health professional for a specific diagnosis compared with 50% of the males who were exclusively heterosexual (p = .049).

DISCUSSION

Among the potential limitations of this study was the response rate of 49%. This rate raises the possibility of selection bias: Gay and bisexual psychiatrists may have been less willing to complete the questionnaire because of fears of stigma from homosexuality. However, the percentage of gay and bisexual respondents in this survey was in the broad range of that reported in the U.S. adult population, roughly midway between the results found by Kinsey, Pomeroy, and Martin (1948) and Laumann, Gagnon, Michael, and Michaels (1994).

There may also have been information bias--namely, that respondents may not have disclosed certain types of sensitive information, either by leaving a question blank, or by answering it inaccurately. Thus, on items such as those regarding homosexuality and substance abuse, the response rates may represent minimum estimates for this population.

Another limitation was that the overall number of gay and bisexual respondents was modest, reflecting the fact that they constitute a relatively small proportion of the overall population. Thus, for adequate statistical power, it was necessary to combine all of the nonexclusively heterosexual subjects into one group for the analyses. Clearly, there may be differences within this group (e.g., differences between exclusively homosexual and bisexual respondents), but these could not be addressed in the present study. Surveys sampling larger numbers of gay and lesbian mental-health professionals can further explore this issue. There is also a risk of Type II errors, that is, we might not find differences when they are present. This risk could be reduced by sampling gay and lesbian psychiatrists through other means, such as through the membership of a gay organization (e.g., the Gay and Lesbian Medical Association). However, sampling from a gay association could lead to other forms of selection bias, as those gays and lesbians who choose to participate in such organizations may not be typical of gays and lesbians as a whole. Moreover, it would be difficult to find a similarly matched comparison group of exclusively heterosexual psychiatrists (i.e., a group involved in an organization similar in its goals and activities).

Another potential limitation is the possible effect of multiple comparisons. The significance levels presented are not corrected for the effects of multiple comparisons (approximately 30). Thus, chance associations may have occurred due to the number of comparisons, particularly when we assessed comparisons of marginal statistical significance. However, classical Bonferroni corrections often tend to be overly conservative in that they assume that all of the comparisons are independent of one another. Because this was not the case in our analyses, we presented the findings without correction, allowing readers to judge the validity of the findings.

Another potential limitation is the possibility of confounding variables. For example, exclusively heterosexual psychiatrists were slightly older than psychiatrists in the other groups. Although this difference in age did not reach statistical significance, age or some other confounding variable may have contributed to the differences observed in this investigation. Again, a study using a larger sample size could effectively examine the contribution of such possible confounding variables.

Despite these limitations, this study offers some basis for speculation concerning the relationship between sexual orientation, the decision to become a psychiatrist, and subsequent professional attitudes and practices. Moreover, as a first study of this population, the data point to differences that can be explored in depth in subsequent research.

Until the 1970s, American psychiatry viewed homosexuality as a disorder listed in DSM-II (American Psychiatric Association, 1968), and consequently gay men and lesbians may have avoided entering the profession. This study suggests that gay men and lesbians have entered the field in the last few decades in a proportion consistent with estimates of their representation in the North American adult population. Moreover, psychiatrists who were not exclusively heterosexual did not differ significantly in most of their approaches toward psychiatric disorders and treatments, with the exception of prescribing medication to their patients. There were no differences between the two groups in biological versus psychological views. This finding suggests that nonexclusively heterosexual psychiatrists represent an integral part of the profession as a whole.

Homosexuality has long been stigmatized and thus can lead to added stress for homosexuals. This stress may account for the increased rate of recreational drug use and the increased tendency to consult a mental-health provider with a diagnosis. High rates of substance use among gay men, lesbians, and bisexuals have been reported elsewhere (McKiman & Peterson, 1989; Skinner, 1994), though without regard to occupation. Conversely, it might be argued that exclusively heterosexual psychiatrists are less likely to use recreational drugs because they have an overall tendency to engage in more socially desirable behaviors.

It is of interest that the nonexclusively heterosexual psychiatrists were more likely to report having a clinical practice in which they did not use medications. This finding suggests that nonexclusively heterosexual physicians may be more likely to have cases involving psychotherapy alone. In particular, gay, lesbian, and bisexual psychiatrists may see more patients concerning sexual orientation issues for which psychotherapy alone is indicated. These psychiatrists, given their own personal experiences of difference, may also be more interested in psychotherapy-only cases in general.

Other findings also suggest that the nonexclusively heterosexual psychiatrists are more humanistic: They prefer psychotherapy-only practices, are less likely to conduct research after residency, and are more open to new experiences (e.g., drug and hallucinogen use). Sexual orientation might thus be part of a phenomenon related to a dimension of heterodoxy-orthodoxy. The finding that exclusively heterosexual psychiatrists were more likely to be Jewish also suggests that these straight psychiatrists may represent an older population of traditionally orthodox Freudians (though age alone as measured in years and in ten-year intervals was not found to be significant).

The issue of possible differences in treatment modalities is important not only because straight psychiatrists were less likely to engage in psychotherapy only, but also because of concerns in the gay and lesbian community over essentialist versus nonessentialist explanations of sexual orientation. A psychiatrist adopting a nonessentialist (i,e., nonbiological) interpretation of the origins of sexual orientation and partner preference may also adopt nonbiological interpretations of other behaviors, including those deemed problematic. Although no differences were found between exclusively heterosexual psychiatrists and other psychiatrists in their views of the origins of psychiatric problems, that straight psychiatrists were less likely to use psychotherapy alone suggests that, at least in terms of some aspects of treatment, an essentialist approach may be favored by straight psychiatrists. Being gay or lesbian may thus affect how psychiatrists approach their patients' problems, specifically favoring more humanistic or nonessentialist treatments.

A provocative hypothesis suggested by these data is that psychiatrists enter the profession because of a specific aspect of their personal history, such as feeling different or marginalized. For example, being Jewish or gay or having first-degree relatives with mental illness may prompt an individual to choose psychiatry as a profession. This hypothesis might explain, for example, the unexpected significant negative associations between sexual orientation and being Jewish or having psychiatrically ill first-degree relatives. Possessing any one of these three marginalizing personal characteristics may be sufficient to prompt an interest in psychiatry. Thus, a gay individual might be less likely to require either of the other two attributes in order to be attracted to a career in academic psychiatry, hence producing the significant negative associations.

Future studies could compare straight and gay physicians of other specialties as well, which has not yet been done. However, examining psychiatrists is particularly important, given that psychiatry had long considered homosexuality a treatable illness and that the etiology of homosexuality, like psychiatric problems, is viewed from conflicting biological and psychological perspectives. The hypotheses tested here and others suggested by our data deserve testing in the future with larger samples of gay, lesbian, and bisexual mental-health professionals.

REFERENCES

American Psychiatric Association. (1968). Diagnostic and statistical manual of mental disorders (2nd ed.). Washington, DC: American Psychiatric Association.

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed.).Washington, DC: American Psychiatric Association.

Armor, D. J., & Klerman, G. L. (1968). Psychiatric treatment orientation and professional ideology. Journal of Health and Social Behavior, 9, 243-255.

Bodkin, J. A., Klitzman, R. L., & Pope, H. G. (1995). Treatment orientation and associated characteristics of North American academic psychiatrists. Journal of Nervous and Mental Disorders, 183, 729-735.

Gladue, B. A., Green, R., & Hellman, R. E. (1984). Neuroendocrine response to estrogen and sexual orientation. Science, 225, 1496-1498.

Hamer, D. H., Hu, S., Magnuson, V. L., Hu, N., & Pattatucci, A.M.L. (1993). A linkage between DNA markers on the X chromosome and male sexual orientation. Science, 261, 321-327.

Isay, R. (1989). Being homosexual. New York: Farrar, Straus, Giroux.

Isay, R. (1995). Becoming gay. New York: Pantheon.

Kinsey, A., Pomeroy, W., & Martin, C. (1948). Sexual behavior in the human male. Philadelphia: WB Saunders Co.

Laumann, E. O., Gagnon. J. H., Michael, R., & Michaels, S. (1994). The social organization of sexuality. Chicago: University of Chicago Press.

Levay, S. (1991). A difference in hypothalamic structure between heterosexual and homosexual men. Science, 253, 1034-1037.

McKirnan, D., & Peterson, P. (1989). Alcohol and drug use among homosexual men and women: Epidemiology and population characteristics. Addictive Behaviors, 14, 545-553.

Monk, M. A., & Thomas, C. B. (1973). Personal and social factors related to medical specialty practice. Hopkins Medical Journal, 133. 19-29.

Pillard, R. C., & Bailey, J. M. (1995). A biological perspective on sexual orientation. Clinical Sexuality, 18, 71-84.

Skinner, W. (1994). The prevalence and demographic predictors of illicit and licit drug use among lesbians and gay men. American Journal ((Public Health, 84, 1307-1310.

Manuscript accepted November 13, 1997

This work was supported in part by grants from The Aaron Diamond Foundation; and the NIMH through Grant P50-MH 43520 to the HIV Center for Clinical and Behavioral Studies, and through Grant K08-MH01420-01.

Direct correspondence to Robert Klitzman, M.D., Department of Psychiatry, Columbia University, Unit #10, 722 West 168th Street, New York, NY 10032; e-mail: [e-mail]