Puberty Blockers

All electrons have spin, and electron spin is a binary quantity: 50% of electrons have spin +¹/₂ and 50% have spin –¹/₂. If an electron was discovered that had a spin of +³/₂ or –³/₂, we would need to rethink our understanding of atomic structure. Such is the degree of confidence that we have in quantum theory, that one would be considered mad to embark on a hunt for spin ³/₂ electrons. However, biological systems are intrinsically heterogeneous. After I secured my first lectureship, I was fortunate to be able to appoint two outstanding physical chemists to work in my research group. We began doing research into the attachment of tissue cells to artificial materials. However, my talented researchers found the experience to be incredibly frustrating: “no two experiments are ever the same”, they told me, whereas in the sort of physical chemistry we were all used to, experimental results were typically reproduced with high precision.

Biologists understand that experiments involving cells or human beings are very complex. No two individuals are the same; experiments yield distributions of results and a great deal of effort goes into analysing them. Statistical tests are used (e.g. the student’s t-test) to quantify significance: two numbers recorded in measurements of some property for two different cell types may be different, but is the difference statistically significant? These tests are based on the assumption that measurements produce a distribution of outcomes, and the significance of the difference between two measurements must be evaluated relative to the distribution of which those measurements are members.

News media like to base stories around individuals; they feel that it helps people to connect with the issues. Kiera Bell features regularly in BBC news pieces about trans issues. Bell began a course of puberty blockers as a teenager and underwent a female-to-male surgical transition as a young adult, but she subsequently decided to detransition. Bell undoubtedly suffered greatly, and in what follows I do not wish to belittle her clearly genuine anguish. However, because of the prominence with which she features in the BBC’s coverage of trans issues, and the prominence of debates about puberty blockers in the media, it is legitimate to consider whether her case is genuinely representative of therapeutic outcomes in the treatment of gender-questioning young people.

The last government commissioned Hillary Cass, a leading paediatrician, to conduct a review of care for gender-questioning young people. Her report is quite a complex and heterogeneous document, blending detailed analyses of medical literature with opinion that is often unsupported by evidence. One of the things that Cass tried to do in her report was to collate numerical data on the treatment of gender-questioning young people. These include data on endocrine (hormonal) practice (p. 168). Of 3,306 patients who had had at least two appointments with the UK Gender Identity Service (GIDS), Cass reported that 27% were referred to endocrinology (893). Of these, 81.5% (727) received puberty blockers, and 64% of these young people went on to receive both puberty blockers and also masculinising/feminising hormones at point of discharge from GIDS. Of the 893 individuals referred to endocrinology, the number who subsequently detransitioned was estimated to be less than 10 (i.e. less than 1%). Thus the likelihood of a young person who is referred to endocrinology undergoing a transition and then deciding to detransition is no greater than, for example, the risk of developing breast cancer due to the oral contraceptive pill, and probably a great deal less.

Overall, the risk of a young person being referred to the GIDS, transitioning and then detransitioning is less than 0.3%. Thus Kiera Bell is representative of a highly unlikely outcome of a referral to the GIDS, and of a slightly larger but still highly unlikely outcome of subsequent referral to endocrinology services. This raises an important question. Why does the BBC repeatedly remind us about Kiera Bell – as though she is in some way representative of a significant sub-set of referrals to the GIDS – when she represents such a tiny fraction of clinical outcomes? The impression that is created, reflected in the angry debates about the Government’s initial approval of a clinical trial of puberty blockers (in line with recommendations in the Cass report), is that puberty blockers are a kind of gateway drug that leads inevitably to surgical transition, detransition and regret. However, Cass’s data do not support that interpretation, although they are undoubtedly complex.

Recognising the complexity of treatments for gender-questioning young people, their distress and the lack of understanding of the origins of gender dysphoria, Hillary Cass repeatedly emphasised the importance of gathering reliable quantitative data and of analysing them using rigorous methodology.

Science is not built on anecdotes

In science, the we eschew anecdotes in favour of empirical data. An anecdote is an account of a singular incident that is not repeatable, whereas empirical data are the results of repeatable experiments and come with an estimate of their precision. Biological systems are incredibly complex and tend to produce distributions of outcomes. There are side-effects to most pharmaceutical products; making decisions about their use involves understanding of the benefits conferred through correct use, and the probability of harm resulting during treatment. Reliable decisions about the adoption of therapeutic approaches, whether for gender dysphoria, cancer or infections disease, can only be made through the rigorous statistical analysis of data sets assembled from carefully constructed clinical trials.

Early in the Covid-19 pandemic, a story emerged about a French boy who had contracted covid, but despite having encounters with 170 people, appeared not to have transmitted his infection to any of them. When tested, the boy was found to have comparatively low levels of infection. There followed a great deal of speculation that this story might provide evidence that children could not spread Covid-19. Many of those speculating – including many professional scientists, often articulating quite sophisticated scientific hypotheses capable of explaining the low probability of transmission of Covid by this boy – were failing to remember something rather fundamental. In epidemiology, one person is not a scientific data point, but an anecdote – and science is not built on anecdotes. Perhaps the boy had a very low susceptibility to the disease for natural reasons (for example, his genetic make-up); in any case, the “attack-rate” of the wild-type Covid virus turned out to be smaller than most people thought at the time. Thus the boy may not have been quite so far from the centre of the normal distribution as one might have expected. Within a year, there was abundant evidence that far from not spreading Covid, children were in fact the main drivers of infection.

No quantity is meaningful without a statement of both the associated units and the experimental uncertainty. In biology, the results of experiments typically follow distributions that must be analysed using statistical tools to evaluate their significance. In grappling with the complex challenges that face us in our everyday lives, science can help us to make better informed decisions, but only if we think about the data in a rigorous, objective fashion that takes account of the uncertainties intrinsic to any experiment.