According to the largest ever study on the genetics of sexual behavior, those who have same-sex partners are more likely to have specific DNA markers.
Researchers say it is possible that thousands of genes, each with a small effect, can influence sexual preference.
Published today in Science, the paper is perhaps the most solid data to date linking genetics to same-sex behaviors.
The researchers emphasize the markers may be unreliable in predicting sexual behavior.
However, the findings provide evidence that some genes do influence the likelihood of same-sex attractions and relationship choices.
Examining genetic markers and survey data about sexual behavior from almost 500,000 people, the researcher identified five new genetic variants not previously associated with homosexuality.
These variants were most common in those who reported being in at least one same-sex relationship.
Two of the markers were shared between men and women; two were exclusive to males and one exclusive to females.
However, the newly identified markers accounted for less than 1% of same-sex behavior.
One of the variants identified in males was associated with male patent baldness, suggesting a tie to testosterone.
When all of the variants were combined across the genome, researchers estimated genetics can account for between 8-25% of non-heterosexual behavior.
Other contributing factors may be environmental influences, in-utero hormonal exposure or social influence later in life.
Previous research linked these biomarkers with a person’s openness to new experiences and increased risks of depression.
Researchers say the findings are limited because people who had a single same-sex experience were counted as non-heterosexual.
They caution experimenting with a same-sex partner may reflect a person’s openness to new experiences rather than actual sexual orientation. Also noted is the data from the UK Biobank did not include details on sexual attraction, just sexual behavior.
Examining genetic markers and survey data about sexual behavior from almost 500,000 people, the researcher identified five new genetic variants not previously associated with homosexuality. The image is in the public domain.
The main limitation of the study is that the focus was on sexual behavior, not orientation or identity.
Additionally, the cohort group consisted of only those with European ancestry. As such, it is not known whether similar results would be seen across different ethnic groups.
In conclusion, researchers say there is no one single “gay gene”, but numerous genes contributing to sexual behavior.
The study was conducted by an international group of researchers. The data was provided by the UK Biobank and 23andMe.
One of the more controversial questions in the neurobiology of human behavior relates to the mechanisms of sexual orientation.
Sexual orientation refers to sexual attraction toward persons of the opposite sex or the same sex.
Just like other behaviors, sexual orientation can be viewed as interplay between specific cerebral processes.
These processes encompass at least three levels:
(1) perception (feeling of attraction triggered by sensory perception),
(2) self-identity (feeling that attraction is related to ‘self’), and
(3) conscious action towards the desired sex.
Most of the hitherto performed studies related to sexual orientation focused on the first level, using various brain imaging methods, and investigating cerebral activation during sexual arousal, elicited by passive viewing of film clips.
A majority of these brain imaging studies of cue-induced sexual arousal seem to agree that the cerebral response is invariant to the preferred sex [1–3] and is primarily related to whether the stimulus is from the desired or non-desired sex, although there are also some exceptions [3–6].
In a series of studies of subjects smelling putative pheromones, we also noticed that activation of the anterior hypothalamus in HoM was reciprocal to that of HeM and similar to the activation pattern of HeW [7,8].
Furthermore, recently, Zhou et al. found that smelling a putative male pheromone enhanced the visual perception of male figures in HoM and HeW, but not in HeM, indicating that hypothalamic activation by the male putative pheromone in HoM and HeW had downstream effects on visual perception with a potential impact on selection of sexual partner [9–11].
While intriguing, these studies only imaged perceptional processes, and could, merely reflect learned behavior.
However, by indicating a link to structures located along the antero-posterior axis of the brain, several of which have been reported as sexually dimorphic (the thalamus, hypothalamus, amygdala), these studies raised important new questions.
One is whether MRI methodology could be used to investigate whether there are any structural brain differences between homo and heterosexual persons.
Another important issue, which has not been emphasized in earlier studies, is whether such differences, if existing, reflect a generally ‘sex-atypical’ sexual dimorphism, or, if they are restricted to areas processing sexual cues and arousal, and, thus, confined to the networks mediating sexual behavior.
These regions vary in different studies, but do, according to a recent meta analysis primarily include the thalamus, hypothalamus, the pregenual anterior cingulate cortex (pACC), and the perirhinal cortex .
Several early post mortem studies suggest structural/histological differences between homosexual men (HoM) and heterosexual men (HeM) along the cerebral midline.
Zhou et al., reported that the size of vasopressin neurons was smaller in the suprachiasmatic nucleus of the hypothalamus in HoM compared to HeM LeVay found that the size of the third interstitial nucleus of the anterior hypothalamus was smaller in HoM than HeM, a finding that was later criticized , and Allen and Gorski observed a larger cross-sectional area of the anterior commissure  among HoM.
Subsequent in vivo brain imaging studies found that the isthmus of the corpus callosum was larger in HoM than HeM , and that HoM and heterosexual women (HoW) had a sex-reversed pattern of hemispheric volume asymmetry .
More recently, Hu et al. reported increased homogeneity in resting state brain activity (suggested to reflect local functional connections) in mid frontal lobe and decreased homogeneity in the middle and inferior occipital lobe .
While congruently suggesting a less pronounced or atypical sexual differentiation of cerebral midline structures, most of these studies investigated a single structure, or used a single metric, and none really addressed the question as to whether the detected differences between homo- and hetero-sexual persons reflected a widespread atypical cerebral sex dimorphism, including several different facets of brain structure, among HeM or restrictive changes around the cerebral midline.
Furthermore, to the best of our knowledge, none of these studies combined structural and functional measures to investigate possible coordinated changes involved in encoding sexual preference which would suggest that related neuroanatomical variations are not merely epiphenomena.
Several previous reports also relayed on mere comparisons between homo-and heterosexual men, limiting the discussions about sex dimorphism.
In the present study, which is part of a larger effort to elucidate the possible neurobiology of sexual orientation and gender identity, structural and resting state functional MRI have been utilized in tandem to investigate possible cerebral correlates to male homosexuality.
We specifically asked ourselves whether the brains of homo and heterosexual men were anatomically different, and how these differences were related to the differences in corresponding measures between HeM and HeW. At variance from our previous investigations  data analysis included both cortical thickness (Cth) and surface area (SA), as each of these metrices have been reported to differ between males and females [17–23].
Cerebral dimorphism with respect to Cth and SA is supposed to be widespread –therefore we used the entire brain as search space.
To further investigate how male sexual orientation was related to cerebral sex dimorphism we also measured volumes of subcortical structures described to differ between men and women–the amygdala, hippocampus, caudate, putamen, and thalamus [17,19,26–29].
The accumbens, which is also of interest in the context of sexual behavior, was not selected as this is a small region difficult to accurately outline on MR images.
In addition, assuming that the direction of sexual attraction is inherently related to perception of self, and, thus, should engage the self-referential neuronal networks [30,31], we investigated functional connections within the so-called default mode network (DMN).
This network is active during rest and mind wandering, and consists of the pregenual cingulate cortex (pACC), the posterior cingulate and precuneus cortex, the right and left inferior, lateral parietal cortex [32–35].
The DMN was of special interest to investigate also because it covers the majority of structures/regions inferred in sexual arousal and behavior (except for the amygdala, hypothalamus and thalamus), allowing us to explore whether possible changes in structures mediating sexual behavior were corresponded by changes in functional connections between these structures.
We also evaluated possible group differences in resting state functional connections from the hypothalamus, and thalamus.
This was done through seed region analyses. These regions were of particular interest as they are reportedly sexually dimorphic, and also involved in sexual arousal.
This study design allowed us to examine:
If HoM differed from HeM analogously to HeW, thus suggesting a ‘female typical’ dimorphism, or, to the contrary?
If such differences were located exclusively within the midbrain networks (preferably those mediating sexual behavior and self-referential processing), but not otherwise.
If HoM differed from both heterosexual control groups.
As expected, the groups differed with respect to the Kinsey scale (p < .0001), but not in age or education (Table 1).
|Unit||HeM (N = 35)||HeW (N = 38)||HoM (N = 30)||F(df)-value||P-value|
|Right D2:D4||ratio||0.97||0.02||1.01||0.02||0.98||0.02||3.678 (2,100);||.038a,b|
|Kinsey scale||0.47||0.83||0.63||0.94||5.70||0.48||418.2 (2,100);||.000c|
a Difference between HeM and HeW, p = .002
b Comparison between HeM and HoM, p = .070
c Comparison between HeM and HoM, and HeW and HoM p < .001
F-values from group comparisons (one way ANOVA). Possible differences in digit ratios between the separate groups were calculated with Scheffe’s post hoc test (p < .05).
We first compared the three groups with respect to Cth and SA. Congruent with previous reports, the parietal lobe cortex (including the postcentral gyrus and the right and left superior parietal lobe), and sensory-motor cortex were thicker in HeW than HeM, whereas the left superior/middle temporal gyrus and the left lateral occipital cortex were thinner (Fig 1, Table 2).
The left superior/middle temporal gyrus in HeW were thinner also compared to HoM, whereas portions of the sensory motor cortex were thicker in HeW than HoM, with no differences between HoM and HeM. However, there was also a deviation from this ‘sex typical pattern’.
First of all, HoM differed from both control groups by having thicker cortex in the right ACC, the superior frontal gyrus and the precuneus (the latter differed only subsignificantly in comparison with HeM), as well as in the left occipito-temporal cortex (covering the extrastriatal body area) (Fig 1, Table 2).
Conjunctional analysis [showing shared clusters from the contrast (HoM–HeW) and (HoM–HeM), p < .05 corrected for multiple comparisons after Monte Carlo permutation] confirmed that HoM differed from both control groups by having thicker cortex in the right ACC-midfrontal cortex and right precuneus (Fig 2), but showed no significant cluster in the occipito-temporal cortex. No other conjunctional clusters were detected. A further finding was that HoM differed from HeM, but not HeW, (thus, showing a ‘female’ pattern), in that their left superior parietal cortex was significantly thicker, and their right cuneus was significantly thinner than in HeM, (Fig 1, Table 2).
Clusters showing significant group difference in cortical thickness.
|Cluster||HeW-HeM (positive -log10(p) values)|
HeM-HeW (negative -log10(p) values)
|HeW-HoM (positive -log10 (p) values)|
HoM-HeW (negative -log10(p) values)
|HeM-HoM (positive -log10 (p) values)|
HoM-HeM (negative -log10(p) values)
|Maximum vertex-wise -log10(p)||Cluster size, cm^2||Talairach Coordinates||Maximum vertex-wise -log10(p)||Cluster size, cm^2||Talairach Coordinates||Maximum vertex-wise -log10(p)||Cluster size, cm^2||Talairach Coordinates|
|R superior + inferior parietal + postcentral gyrus||4.4||14.1||18–46 57||2.5||11.0||5 2 50|
|L superior + middle temporal gyrus||-4.2||14.3||-58–25–7a||-4.5||34.9||-48–25–7b||-3.8||25.3||-55–58 5†|
|L precuneus + superior parietal cortex||2.5||10.0||-8–68 50||-3.4|
|L lateral occipital cortex||-2.6||13.7||-12–67–18||-2.6||11.8||-48–64 1|
|R superior and rostral frontal cortex||-4.8||60.4||14 39 16c||-3.8||14.9||15 33 17|
|R cuneus, pericalcarine cortex||3.3||22.2||26–87–8||12.6||4–71 10|
Statistical threshold is p < .05, corrected for multiple comparisons (according to Monte Carlo permutations). Demeaned age was used as a nuisance covariate. The filter was 10 mm. The Talairach’s coordinates indicate location of maximum difference, the ‘Region’ column describes the coverage of the respective cluster. Italics indicate clusters calculated at p < .05 uncorrected.
R = right; L = left
a covers the L temporo-occipital cortex
b covers also the lateral occipital lobe
c covers also the precuneus
Post hoc analyses: Cortico-cortical covariations in Cortical thickness
Several recent studies have found that cerebral functional networks have an intrinsically cohesive modular structure where the modules are composed of functionally as well as anatomically related brain regions, and these networks can be identified by maps of covariance .
To investigate whether the observed differences between HoM and the controls were interrelated, post hoc cortico-cortical covariation analyses were performed using the two significant clusters detected in conjunctional analyses (right midfrontal-cingulate cluster and right precuneus cluster, Fig 2) as seed regions (p < .05 after Monte Carlo correction for multiple comparisons), as previously described .
In short, the mean Cth from each region of interest (ROI) was first extracted in each subject.
Next, the data from each seed ROI was used as covariate of interest to test possible group difference in the covariation pattern between the respective seed region and the rest of the brain.
Possible differences in covariation patterns were tested between HoM and HeM, HoM and HeW, and between HeW and HeM (qdec statistics within the FreeSurfer software, p < .05 after Monte Carlo correction for repeated comparisons). HoM showed a significantly stronger correlation compared to both HeM and HeW between the Cth in the precuneus seed and the anterior and mid cingulate, the occipital cortex, the inferior frontal and insular cortex, S1 Fig, S1 Supplemental Information). No difference was detected between HeM and HeW, and no group differences were found with respect to the covariation pattern from the midfrontal-cingulate seed.
Post hoc analyses: Correlation between Kinsey scores and cortical thickness
To assess whether the differences in Cth between homo and heterosexual men were specifically linked to sexual orientation, we carried out post hoc analyses including all three study groups (HoM, HeM, and HeW), and testing whether there was any significant correlation between Cth and Kinsey scores. Using the qdec statistics, Kinsey scores were employed as the covariate of interest, (age was a nuisance covariate) in relation to Cth in each vertex of the brain (p < .05 with Monte Carlo correction for multiple comparisons). The analysis showed a significant positive correlation between Kinsey scores and the Cth in the occipito-parietal and precuneus cortex (S2 Fig, S1Supplemental Information).
In addition, there was a sub significant (p < .05 uncorrected) cluster indicating a positive correlation between Kinsey scores and Cth of the ACC. No clusters indicating inverse correlations were detected.
Press Office – AAAS
The image is in the public domain.
Original Research: Open access
“Large-scale GWAS reveals insights into the genetic architecture of same-sex sexual behavior”.Andrea Ganna et al.