Several documents, including the Endocrine Society Guidelines, refer to the diagnostic importance of adolescents starting puberty. There is a belief that some youth cease to be transgender/dysphoric during the early stages of puberty and that ‘persistence’ can only be assessed after youth have experienced that first stage of natal puberty.

Here’s the Endocrine Guidelines 2017

“At the present time, clinical experience suggests that persistence of GD/gender incongruence can only be reliably assessed after the first signs of puberty.”

“Adolescents with GD/gender incongruence should experience the first changes of their endogenous spontaneous puberty, because their emotional reaction to these first physical changes has diagnostic value in establishing the persistence of GD/gender incongruence.”

The belief in the diagnostic importance of the early stages of puberty, leads in turn to the requirement that adolescents go through the first stage of a natal puberty, and the clinicians only prescribe blockers once they have noted increased distress at those early pubertal changes. The Endocrine Society’s criteria for prescription of blockers includes the requirements:

“gender dysphoria worsened with the onset of puberty”.

As mentioned in our earlier blog, the Endocrine Guidelines do not include any reference in support of this recommendation.

However, this quote does appear to correspond to a journal article by Steensma et. al., 2011.

Steensma TD, Biemond R, de Boer F, Cohen-Kettenis PT. Desisting and persisting gender dysphoria after childhood: a qualitative follow-up study. Clin Child Psychol Psychiatry. 2011;16(4):499–516.

We will therefore look at this in detail.

The abstract for this paper makes a wide number of claims:

“Adolescents with persisting gender dysphoria (persisters) and those in whom the gender dysphoria remitted (desisters) indicated that they considered the period between 10 and 13 years of age to be crucial. They reported that in this period they became increasingly aware of the persistence or desistence of their childhood gender dysphoria. Both persisters and desisters stated that the changes in their social environment, the anticipated and actual feminization or masculinization of their bodies, and the first experiences of falling in love and sexual attraction had influenced their gender related interests and behaviour, feelings of gender discomfort and gender identification.”

It offers as a conclusion:

Based on the significance most adolescents attribute to the period between the ages of 10 and 13, we suggest that clinicians should concentrate clearly on what happens in this phase of develop­ment. It is recommended to specifically address the adolescents’ feelings regarding the factors that repeatedly came up as relevant in our interviews (i.e. the effects of the changing social environ­ment, the response to anticipated or actual puberty, and the emerging romantic/sexual feelings and sexual partner choice), before any medical steps are taken (e.g., to suppress further pubertal development).

Steensma et al.’s above conclusion makes some very specific recommendations that have influenced  clinical practice in settings, including the UK Gender Identity Service for Children.

Here is the ‘Evidence‘ section of the UK NHS Children’s Gender Identity Service website:

“Young people indicated that the period from 10 and 13 years to be most crucial in their feelings related to gender dysphoria. For both, the young people who continued having these feelings and for those where the feelings of gender dysphoria became less prominent, three main factors seem to have had an impact on their gender identity development. Firstly, the changes in social environment (gender roles and expectations become more distinct during this period of their life); secondly, the effects of a changing body through puberty; and thirdly the experience of romantic feelings and falling in love (Steensma et al, 2011).”

Other papers by Steensma have been criticised for poor and inaccurate research analysis, for drawing conclusions that are not based on the findings, and for overstating conclusions beyond what the data can support.

Key claims to consider as we look into the actual data in the Steensma paper are:

1. Is there robust evidence that “emotional reaction to these first physical changes has diagnostic value in establishing the persistence of GD/gender incongruence”.
2. Is there robust evidence that “the effects of the changing social environ­ment, the response to anticipated or actual puberty, and the emerging romantic/sexual feelings and sexual partner choice create changes in gender identity at this point in time”
3. Is there robust evidence that “”tanner stage 2 is a diagnostically important period for gender identity?” (Steensma’s conclusion does not mention tanner stage, instead mentioning age 10-13, but clinical guidelines now focus on tanner stage rather than age)
4. Is there evidence that “a significant number of trans children, previously undistinguishable pre –puberty, desist in the first stage of puberty” (between tanner 1 and 2)?

To the data! (Steensma et. al. 2011)

The study selected a sample of 25 adolescents aged 14-18, all of whom had been registered with the Dutch Gender clinic in childhood, but only 14 of whom had been still registered with the service at age 12-14. The 14 who had been with the service at age 12-14 (and were still with the service) were considered the ‘persisters’ (7 ‘male’ 7 ‘female’ – amab/afab?). The 11 who had left the service some time before the age of 12-14 were considered the ‘desisters’ (6 ‘male’ 5 ‘female’ – amab/afab?). The persisters had applied for puberty blockers. The desisters had not applied for puberty blockers.

The Steensma study interviews this sample at ages 14-18 and asks a wide range of qualitative questions, asking them to reflect upon their earlier views and experiences.

All 25 children had been diagnosed during childhood with Gender Identity Disorder (DSM 4). The diagnosis of Gender Identity Disorder is no longer used, and has been heavily criticised for not distinguishing between children who are transgender and those who are simply non-conforming, with no wish to change their gender and no need for medical interventions. The now discredited statistics on desistance, also published by Steensma amongst others, are known to be flawed as they relied upon the DSM 4 diagnosis, counting non-conforming children alongside transgender children.

With a DSM4 diagnosis, we cannot know how many of the original sample of 25 were just gender non-conforming. The possibility that a large number of children in this sample of 25 were non-conforming rather than transgender is given credence by the fact that the paper refers throughout to issues that are not centred on identity – the paper focuses predominantly on descriptions of gendered interests, play preferences and gender expression (as opposed to on identity).

It is also interesting to note that none of the 25 children in this sample had socially transitioned before the age of 12.

The paper makes the claim that ‘Adolescents with GD/gender incongruence should experience the first changes of their endogenous spontaneous puberty, because their emotional reaction to these first physical changes has diagnostic value in establishing the persistence of GD/gender incongruence’.

As we will demonstrate, this recommendation is built upon woefully shaky foundations:

Steensma et al (2011) claim to have identified three diagnostic areas:

1. social divisions
2. pubertal physical changes and
3. sexual orientation

all of which they claim are diagnostically critical in distinguishing between ‘persisters’ and ‘desisters’ between the ages of 10-13 years old.

The paper outlines the different ways that ‘persisters’ (those who were in the system at age 12-14 and applied for puberty blockers) and ‘desisters’ (those who had left the system before age 12 and never applied for puberty blockers) remember feeling ‘at around age 10-13’

1. Social Divisions

The ‘persisters’ recall social divisions between boys and girls increasing ‘at around age 10’. As the divisions between boys and girls increased, so did their wish to be grouped with the ‘other’ gender (with the gender matching their identity?) and they increased in their wish to socially transition. The ‘desisters’ did not respond in the same way. At this period of increased social divisions between boys and girls, the ‘desisters’ were not troubled to be grouped with their natal gender, and did not wish to socially transition.

Steensma et al interpret the above as evidence that the period of increased social division at around age 10 is diagnostically important. That we need to wait until social divisions between boys and girls increase at around age 10, and see whether children wish to be grouped with their natal gender without wish for social transition (=desisters) or whether they wish to be grouped with children of the opposite gender to their assumed gender and wish to socially transition (=persisters).

An alternative possibility (which cannot be determined from this data) is that the desisters were always gender non-conforming children, and the persisters were always transgender children. An alternative interpretation of the same research data is that when social divisions increase, the transgender children wish to be grouped with their identified gender, whereas the non-conforming children are untroubled with being grouped with their natal (=identified) gender. With this interpretation, the age of 10 and the responses to increased social division at this age, is not in fact diagnostically important. Instead, to distinguish transgender children from non-conforming children we need to ask them about their identity. Something, from the data presented, this study does not do.

This ‘finding’ focuses on an age, in this case the age of 10 (a time where according to Steensma et al, gender divisions increase in the Netherlands). This ‘finding’ is not linked to any stage of pubertal development or to any tanner stage.

2. View of puberty

The second distinction proposed between the ‘persisters’ and ‘desisters’ relates to their reported ‘view of puberty’. When interviewed several years later (at age 14-18) the ‘persisters’ recall having been very distressed by puberty:

“When I was 13, I started to menstruate and my breasts started to grow. I hated it! If we would have had a train station in our town I would definitely have jumped in front of a train. I didn’t go to school anymore, lost my friends and became totally withdrawn”.

The desisters, being interviewed at ages 14-18, do not recall being distressed about the physical changes of puberty. The desisters were almost by definition not distressed by puberty as they are the group who had left the service before age 12 and had not applied for puberty blockers. As throughout this paper, tanner stage of puberty is not discussed.

Steensma et al interpret this as evidence that experiencing “the first changes of their endogenous spontaneous puberty” was a critical diagnostic stage.

An alternative interpretation of this same research is that children who are distressed about puberty, who are still in the gender service at age 12, and who apply for hormone blockers, can be considered as ‘persisters’ (transgender). Children who are not distressed about puberty, are no longer in the service at age 12, and do not apply for hormone blockers, are ‘desisters’ (not transgender).

There is no evidence in this study that the desisters were distressed about the idea of puberty beforehand and desisted at age 11 (remember they had all left the gender service before age 12). There is no reference to tanner stage 2 in this study, and no evidence at all that desisters were distressed at tanner 1 but un-distressed at tanner stage 2.

The study even makes it explicitly clear that the desisters were not even distressed at the idea of puberty:

“For the desisters the anticipated feminization or masculinization of their bodies was not explicitly reported as particularly distressful”.

The desisters were not even distressed about anticipated puberty.

The most logical interpretation of this data that we can make is that we should not give puberty blockers to youth who have not applied for puberty blockers. And we should not give puberty blockers to youth who have left the gender service before age 12. It is our conclusion from the presented data that assuming the children left the service of their own accord, rather than leaving for other reasons, such as denial or delay in treatment, it seems most likely that children who have left the service before age 12 and have not asked for puberty blockers are probably not in need of puberty blockers.

Significantly, and at odds with the paper’s abstract, recommendations and conclusions, this study provides no evidence that children who are distressed about anticipated puberty desist after undergoing the first stages of puberty.  It also fails to make a single reference to tanner stages.

Yet this is considered evidence that “Adolescents with GD/gender incongruence should experience the first changes of their endogenous spontaneous puberty”. The study provides no evidence to back up that conclusion.

3. Sexual attraction

The third section which Steensma et al. find to have diagnostic importance is that of developing sexual attraction. In the study (interviewing 14-18 year olds), all 14 ‘persisters’ described themselves as attracted to their natal sex, and saw themselves as heterosexual (transgender and heterosexual).

Steensma et al. extrapolate from this a conclusion that the development of sexual orientation is, in and of itself, diagnostic – that youths who are attracted to their natal gender but consider themselves heterosexual are therefore transgender.

This finding is problematic on multiple levels.

Firstly, and inexplicably, Steensma et al. consider it a noteworthy research finding that a transgender person attracted to their natal gender considers themselves heterosexual.

Secondly the sexual orientation of a sample of 14 transgender individuals (all describing themselves as heterosexual) cannot be considered diagnostic as even from a cursory review of the literature it is clear that transgender people can have a range of sexual orientations.

Thirdly, whether all 14 are heterosexual cannot itself be relied on – remember these are interviews with 14 – 18 year olds, adolescents at ages where many cisgender youth may not be open about their sexuality, why then should transgender youth be any different. Moreover these are transgender youths who are reliant upon a service for medical interventions and it can be speculated are attempting to provide the ‘desired answers’ to navigate their way through in a Gender Identity Service ruled by hetero/cis-normative gate keepers.

Also many of the ‘persisters’ seem to emphasise ambivalence or reluctance to start dating “I just don’t want to date now”.

There is a clear hetero cis normative bias, particularly for assigned females.  The Steensma et al (2011) analysis notes that the desisting girls were all cisgender heterosexual and suggests that this is also diagnostic (implicit assumption that cisgender lesbians do not exist). The desisting boys it notes expressed a variety of sexual orientations.

Even within this data set it is unclear how Steensma et al consider sexuality diagnostic. Note again, there is nothing here about tanner stage, nor about at which stage of puberty sexual orientation became clear.

As an aside that I won’t go into here, the study is also methodologically flawed (it describes itself as applying grounded theory, but omits to include key parameters that are critical for a grounded theory approach).

Summary

What the Steensma et al. 2011 study claims to prove:

“Adolescents with persisting gender dysphoria (persisters) and those in whom the gender dysphoria remitted (desisters) indicated that they considered the period between 10 and 13 years of age to be crucial. They reported that in this period they became increasingly aware of the persistence or desistence of their childhood gender dysphoria. Both persisters and desisters stated that the changes in their social environment, the anticipated and actual feminization or masculinization of their bodies, and the first experiences of falling in love and sexual attraction had influenced their gender related interests and behaviour, feelings of gender discomfort and gender identification.”

What the Steensma et al. 2011 study actually proves:

Social divisions: 25 people aged 14-18 reminisced about their experiences at the age of about 10, a time when social divisions between boys and girls increases in the Netherlands. 14 children who were still registered with the gender identity service in adolescence recall at that time of increased social division having wanted to be grouped with the children of the opposite gender to their assigned gender and wanting to socially transition. 11 children who had left the service before the age of 12, recall having wanted to be grouped with their assigned at birth gender and having not wished to socially transition.

View of puberty: 11 children who were in the service in childhood but left the service before the age of 12 and did not apply for puberty blockers, remember not having been distressed by the idea or the reality of pubertal changes. 14 children who were still in the service in adolescence and who applied for blockers remember being distressed at pubertal changes.

Sexual attraction: In a sample of 25 people registered at the gender clinic in childhood, when interviewed in adolesence, 14 transgender people were heterosexual, 5 cisgender girls were heterosexual and 6 cisgender boys had a variety of sexual orientations.

What policy recommendations this study makes:

Based on the significance most adolescents attribute to the period between the ages of 10 and 13, we suggest that clinicians should concentrate clearly on what happens in this phase of develop­ment. It is recommended to specifically address the adolescents’ feelings regarding the factors that repeatedly came up as relevant in our interviews (i.e. the effects of the changing social environ­ment, the response to anticipated or actual puberty, and the emerging romantic/sexual feelings and sexual partner choice), before any medical steps are taken (e.g., to suppress further pubertal development).

What policy recommendations this study can justifiably make:

We should not give puberty blockers to youth who have left the service before the start of puberty, who are not distressed at the idea of puberty and who have not applied for puberty blockers. We should only give puberty blockers to youth who are still in the service, who are distressed at the idea of puberty and who apply for puberty blockers.

Tanner 2?

It is noteworthy that the Steensma study make no reference to tanner stage 2, instead focusing on the age of around 10-13 and the stage of emerging romantic/sexual feelings. Current practice in many countries has moved on from designating a minimum age for puberty blockers (age 12) to a stage based approach (tanner stage 2).

In other countries the approach is reportedly one where a transgender child’s identity is believed in childhood, where children approaching puberty are reassured that puberty blockers will be available at tanner stage 2 if required, where clinicians proactively monitor tanner stage, and where, at tanner stage 2, if a child is distressed at the idea of pubertal changes, puberty blockers are prescribed in a timely fashion.

In the UK flawed evidence like this Steensma study provides the foundation of an approach which is harmful to transgender children and adolescents. The UK approach uses flawed desistance statistics and studies like this one on the diagnostic importance of puberty to argue that transgender children cannot know their identity until puberty. The UK uses this (plus other flawed research again by Steensma) to argue against early social transition even for insistent, persistent, consistent and deeply distressed transgender children. The UK belief that previously trans children will desist at puberty means that pre-pubertal children are given no reassurance at all that puberty blockers will be available, leading to increased stress as puberty approaches. Parents report that the UK service does not seem to see any urgency in prescribing promptly at tanner 2, with parents feeling the need to fight for the service to monitor developing tanner stage and to prescribe in a timely fashion at tanner 2. Parents are confused about what possible reason clinicians could have for delaying prescription of hormone blockers to distressed children at tanner 2. They don’t understand why there is no sense of urgency or timeliness from the UK Children’s Gender Identity Service.

My guess is that this Steensma et al. 2011 study is part of the reason UK Children’s Gender Identity Service clinicians seem so reluctant to prescribe puberty blockers promptly at tanner stage 2. If they accept Steensma et al.’s conclusions and policy recommendations at face value (which they seem to), then they are led to believe that there is an unknown point in natal puberty where previously insistent trans children will suddenly ‘desist’. That there is no way of knowing beforehand which kids will desist. That the longer they can delay blockers the more likely adolescents are going to experience the elusive (and unknown) point in puberty when something as unpredicatable as falling in love could cause them to ‘desist’. Another child will be saved from transdom! And one child saved from transdom is worth inflicting major emotional harm on those who remain trans.

Summary:

This piece of research over states its conclusions and draws policy recommendations that are not in any way supporting by the data. This publication is cited twice in the new Endocrine guidelines. It is included in the references for WPATH Standards of Care 7. It is cited in numerous articles. It is quoted at length on the website of the UK Children’s Gender Identity Service. The claims made in this study are not robust and must be discarded.