Proc. Nati. Acad. Sci. USA Vol. 89, pp. 7199-7202, August 1992 Neurobiology Sexual orientation and the size of the anterior commissure in the human brain (homosexuality/sex difference/sexual differentiation) LAURA S. ALLEN AND ROGER A. GORSKI Department of Anatomy and Cell Biology, and Laboratory of Neuroendocrinology, University of California, 405 Hilgard Avenue, Los Angeles, CA 90024 Communicated by C. H. Sawyer, April 24, 1992 ABSTRACT The anterior commissure, a fiber tract that is not directly related to reproductive function: there is an larger in its midsagittal area in women than in men, was increase in the frequency of left-handedness (19-21), child- examined in 90 postmortem brains from homosexual men, hood play behavior more often resembles that of girls (22), heterosexual men, and heterosexual women. The midsagittal and scores on tests ofvisuospatial and verbal abilities (23-26) plane of the anterior commissure in homosexual men was 18% as well as cerebral lateralization (27) resemble those of larger than in heterosexual women and 34% larger than in heterosexual women more than those of heterosexual men. heterosexual men. This anatomical difference, which correlates However, others using different criteria have not found a with gender and sexual orientation, may, in part, underlie correlation between sexual orientation and handedness (28) differences in cognitive function and cerebral lateralization or cognitive abilities (29). We hypothesized that in humans, among homosexual men, heterosexual men, and heterosexual as in laboratory animals, there is a correlation in the direction women. Moreover, this rinding ofa difference in a structure not of sexual differentiation within a given individual between known to be related to reproductive functions supports the sexually dimorphic functional characteristics, including sex- hypothesis that factors operating early in development differ- ual orientation, and structural sex differences in the brain that entiate sexually dimorphic structures and functions of the we have identified (11, 12, 16). We now report that a sexually brain, including the anterior commissure and sexual orienta- dimorphic structure not directly related to reproductive func- tion, in a global fashion. tion, the AC, in homosexual men more closely resembles that of heterosexual women than that of heterosexual men in A fundamental question of human behavior is the relative terms of area at the midsagittal plane. contribution of genetic, fetal or neonatal hormonal, and Tissue from presumed homosexual women was not exam- postnatal environmental factors to the etiology of sexual ined in this study because AIDS, which was present in most orientation. In laboratory animals, considerable attention has ofthe homosexual male subjects, has not specifically affected focused on observations that levels of gonadal hormones homosexual women. Therefore, sexual orientation is gener- during a critical period of development determine the sexual ally not noted in hospital medical records of women. In differentiation of the brain in terms of many sexually dimor- contrast, sexual orientation is generally specified in medical phic functions and structures (1). In contrast, little is known autopsy records of all men with AIDS or other diseases that regarding sexual differentiation of the human brain. In terms can be either sexually transmitted or result from immunode- of function, humans exhibit sex differences in reproductive ficiency. behavior, gonadotropin secretion (2-4), childhood play be- havior (5), and cognitive abilities and cerebral lateralization METHODS (6-8). In regions of the brain presumably involved in repro- ductive function, there are reports of sex differences in the The brains used in this study, which were obtained from three suprachiasmatic nucleus (9), the interstitial nuclei of the Southern California hospitals between 1983 and 1991, had anterior hypothalamus 1-3 (10, 11), and the bed nucleus ofthe been removed within 24 hr postmortem and placed directly stria terminalis (12). In regions of the brain not directly into acetate-buffered 10%6 formalin. The present investigators related to reproductive function sex differences occur in collected 256 samples of brain tissue containing the midsag- cerebral asymmetry (13), in the shape of the corpus callosum ittal region ofthe AC from brains that appeared unaffected by (CC) (14, 15), and in the midsagittal area of the massa neuropathology during routine autopsy. Subsequent to his- intermedia and anterior commissure (AC) (16). tological analysis of brain samples, medical records were Several sexually dimorphic functions and structures re- examined for results from histoneuropathology, neurological portedly differ in homosexual people. Studies indicate that disorders, cause of death, age, gender, and sexual orienta- homosexual men exhibit a luteinizing hormone surge that is tion. Subjects were eliminated from this study when medical intermediate between heterosexual men and women (2, 3), records indicated histoneuropathology or disorder(s) poten- although others have failed to confirm this (4). Morphological tially influencing the AC (e.g., demyelinating disease; n = differences have been identified in homosexual men in nuclei 59), bisexual orientation (n = 1), or AIDS when all risk of the brain that are in regions that influence reproductive factors for the disease were denied by the patient (n = 3). physiology and behavior: the volume of the suprachiasmatic Male and female subjects were classified as heterosexual nucleus may be larger than in heterosexual men and women when medical records did not indicate homosexual orienta- (17); and the interstitial nucleus ofthe anterior hypothalamus tion. This procedure generated 34 homosexual men, 84 3, which we found to be larger in men than in women (11), is heterosexual women, and 75 heterosexual men. Before the smaller in homosexual men than heterosexual men (18). evaluation of AC area and from the identification codes that Homosexual males also exhibit differences in characteristics included age, sex, and sexual orientation, the subjects were age-matched as closely as possible to obtain 30 triplets-each including a homosexual man, heterosexual man, and hetero- The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Abbreviations: AC, anterior commissure; CC, corpus callosum. 7199 Downloaded by guest on September 30, 2021 7200 Neurobiology: Allen and Gorski Proc. NatL Acad. Sci. USA 89 (1992) sexual woman (unmatched subjects were not further evalu- 27 ated). 24 Without knowledge of gender or sexual orientation, the E E 21 blocks of tissue containing the AC were cut by hand at the (E midsagittal plane, and this surface was placed against a glass 18 0 * adjacent to a ruler at the same plane, photographed, and U- . 0 15 made into slides. These images were projected onto white paper and traced twice at different times. The area of each 12 outline was determined by using a Bioquant Hipad digitizer 9 program 0 adjusted to correct for magnification (Bioquant IBM 6 version 2.1; R & M Biometrics, Nashville, TN), and the two (I) 0 measurements of each AC were averaged. In an attempt to 3 verify that individuals who died of AIDS did not have neuropathology known to increase gross neural structure- homosexual heterosexual heterosexual namely, edema and vacuolization, seven ACs from homo- men men women sexual males who died of AIDS and seven ACs from hetero- FIG. 1. Midsagittal area of the AC in homosexual men, hetero- not AIDS were sexual subjects who did die of randomly sexual men, and heterosexual women. Data are means + SEMs. selected, sectioned at 60 pum, stained with thionin, and Filled circles represent individual subjects without AIDS, and open examined under a light microscope by a neuropathologist. triangles represent individual subjects who died of AIDS. One-way analyses of variance (ANOVA) were performed between the three groups for age, area of AC, brain weight, and Mann-Whitney test, P < 0.0001). The AC in females was area of AC divided by brain weight, postmortem period 13.4% or 1.4 mm2 larger than that of heterosexual men before autopsy, and period between autopsy and photogra- (Bonferroni t procedure and Mann-Whitney test, P > 0.05). phy. To ascertain that our results were not from the two When the area of the AC in homosexual men without the largest ACs of homosexual males, ANOVA were performed two subjects with a relatively large AC is compared with the excluding these two samples. Similarly, ANOVA were used two heterosexual groups, a significant difference remains in to compare homosexual men with AIDS, homosexual men the area ofthe AC and area adjusted for brain weight between without AIDS, heterosexual men with AIDS, and heterosex- heterosexual men but not heterosexual women (Table 1). ual men without AIDS. Post hoc comparisons with the Clearly, further studies are needed to determine whether the Bonferroni t procedure were used to determine differences two large ACs represent a normal variation in homosexual between the individual groups. Due to the large range of men and whether the AC is, in fact, larger in homosexual men variation in the size of the AC, Mann-Whitney comparisons than heterosexual women. Although the AC size did not of AC area between the three groups were made. Pearson's correlate significantly with brain weight (r = 0.019, P =