December 7, 2017

expert reaction to research on genetic links to male sexual orientation

A preliminary genome-wide association study, published in Scientific Reports, investigates whether there are genetic links to male sexual orientation.

Dr Jeffrey Barrett, Director of ‘Open Targets’ at the Wellcome Trust Sanger Institute, said:

“Like nearly every complex human trait we study, the genetics of sexual orientation will be multifactorial (as the press release notes), meaning that hundreds (or likely thousands) of genetic differences will each contribute in a tiny way to sexual orientation. What this means is that to study traits like this requires looking at the genomes of tens of thousands (or even hundreds of thousands) of people.

“This study is way, way, way too small to draw any meaningful conclusion. None of their findings meets the accepted thresholds for statistical significance in a genome-wide association study (which is why it is published in Scientific Reports). The comments about SLITRK6 and TSHR are utter speculation, and don’t belong anywhere near a modern genetic study — we had decades of such claims that never held up because they didn’t meet statistical significance.”

Prof. Robin Lovell-Badge FMedSci FRS, Group Leader at The Francis Crick Institute, said:

“The topic of this paper is important if we are to learn more about the influence of genes on aspects of our behaviour, but this is one that is notoriously difficult to study. It is mired in politics, problems of securing funds, and difficulties in having sufficient numbers of volunteers for the research. So this is a brave attempt. Genome-Wide Association Studies (GWAS) are also problematic, and are frequently underpowered; meaning that the sample sizes are too small to allow any robust conclusions. This is particularly so for traits like being gay that are likely to involve many genes and where ‘environmental’ influences (perhaps both in utero and postnatally) can have a strong effect. Moreover, correlation does not mean causation. (The increase in value of my flat correlates with my increase in age – but I doubt that there is any causal relationship.) Apart from low numbers of individuals in the study, the authors also point out that they were restricted to men of European descent; again a shortcoming preventing any generalisations. So when the authors spell out the caveats about their research and the findings, these have to be taken seriously.

“The new data on genetic associations they report here does not show sufficient statistical significance to make any formal link between a gene or chromosome region and being gay or heterosexual. However, their data does tend to reinforce some previous findings of associations, which were also relatively weak, with a region of Chromosome 8 and the X chromosome. Regions of chromosome 13 and 14 also appear from the new work to be worth paying attention to in future studies. The authors also suggest some candidate genes – but this is really just speculation and it is impossible to tell how any particular gene might influence any aspect of behaviour without substantially more work – and this can be very challenging in humans where studying gene function in a way that controls for variations in genetic background and environment is very difficult. Moreover, even if a gene variant does show some correlation with sexual orientation, this does not mean that the gene is in any way responsible for being gay – it just means it has some association with a trait that is more likely to found in the relatively few people involved as subjects in the study. This could be better social awareness or being brave enough to acknowledge that they are in a minority.”

Prof. Gil McVean FRS FMedSci, Professor of Statistical Genetics at the University of Oxford, said:

“The researchers have found weak evidence for genetic variation that influences self-reported sexual preferences in men.  However, the sample size is small, the results have not been replicated in an independent study and the level of evidence presented doesn’t meet the threshold of significance typically required within the field.  The press release is appropriate, but I don’t think the work would have been published if it were on a less controversial topic.  It is – at best – preliminary.

“There is limited evidence from previous work for genetic influences on sexual preferences.  This work confirms that if there are genetic effects, they are extremely weak and scattered across the whole genome.  So there is certainly no single gene for influencing sexual preference – rather (like height, mathematical ability or risk of diabetes), many small effects that, even together, play only a weak role in shaping sexual preference.

“The specific biological hypothesis put forward in this paper does not have anything more than highly tenuous / circumstantial evidence.  Rather, as above, sexuality is likely influenced by many different factors, including environment, experience and (likely) some aspects of innate biological variation.

“I can see no major implications of this work or how it could be useful in the future.  The genetic effects are far too weak to be of any predictive or diagnostic value.  All biology – including the origins of sexuality – is interesting at some level, but I see no direct applications of such research.”

* ‘Genome-Wide Association Study of Male Sexual Orientation’ by Alan R. Sanders et al. will be published in Scientific Reports at 2pm UK time on Thursday 7^th December, which is also when the embargo will lift.

Declared interests

None declared.