Marianne Van Der Loos

OF THE MAIN FINDINGS AND GENERAL DISCUSSION

In this chapter the main findings of this thesis are summarized. Subsequently overall strengths and limitations, and clinical implications are provided. Since adolescent transgender healthcare remains controversial, the findings of the studies presented in this thesis will also be discussed within the societal context. Finally, suggestions for further studies are made.

Part 1 – Trajectories of transgender adolescents
In Chapter 2 we examined treatment trajectories in children and adolescents referred to the gender identity clinic of the Amsterdam UMC, location Vrije Universiteit Amsterdam. It was found that the steep increase in recent referrals is mainly generated by pubertal (i.e. ≥ 10 years of age at intake) individuals assigned female at birth (AFAB). The overall median age at intake is higher in AFAB than in individuals assigned male at birth (AMAB). A shift in ratio of sex assigned at birth towards AFAB occurred around 2012. The observed shift in AMAB:AFAB ratio is in line with previous studies. Age at intake gradually increased over time, possibly as an unfortunate result of a long waiting list to enter the gender identity clinic. This only slightly affected the age at which medical treatment, gonadotropin-releasing hormone agonist (GnRHa) and/or gender-affirming hormones (GAH), was started. Almost half of AMAB started GnRHa, versus three-quarters of AFAB. The main reason for this seemed that AMAB were less likely to meet the criteria for a diagnosis of gender dysphoria. This was also the case in a recent study performed in the USA. The proportion of people starting GnRHa has mainly increased over time in individuals first visiting < 10 years of age. Compared with AFAB, AMAB were 1.5 years younger when starting GnRHa. Nearly all AMAB and AFAB on GnRHa went on to start GAH. Age at start of GAH was just slightly higher in AFAB. Keeping in mind the pronounced age difference at start of GnRHa, this means that in general AMAB were on mono GnRHa treatment longer. In the sporadic cases GnRHa treatment was discontinued, the main reason was remission of gender dysphoria. Additionally, a pronounced difference was found between the percentages of persons undergoing gonadectomy before and after July 2014, when a change in law no longer required people to undergo gonadectomy to change their legal sex. Continuation of GAH treatment in people treated according to the Dutch Protocol was assessed in Chapter 3. By merging this particular cohort to both the hospital’s prescription registry and the prescription registry that is maintained nationally by Statistics Netherlands (Centraal Bureau voor de Statistiek; CBS), we found that 98% continued GAH treatment. Discontinuation was not associated with age at first visit, age at start of GnRHa treatment, age at start of gender-affirming hormone treatment, puberty stage at start of GnRHa treatment, gonadectomy, year in which people first visited, or year in which GAH treatment was started. Although the subject of continuation is studied by Roberts et al., who investigated continuation of GAH four years after start of treatment, a similar study in this population has not been reported. The Roberts study found 74.4% of individuals starting GAH before age 18 had continued treatment. However, it does not mention how many of these adolescents were using puberty suppression prior to GAH. Part 2 – Bone health in transgender adolescents BMD development during GnRHa and GAH treatment, and after long-term GAH treatment was assessed in Chapter 4. At the lumbar spine, total hip, and femoral neck BMD Z-scores decreased during GnRHa treatment in both AMAB and AFAB, which has become a common finding. Previous studies assessing BMD Z-scores showed they did not completely catch up after short-term GAH. However, in our study at long-term follow-up a compensatory increase in bone mineral accrual was seen at all three regions in AFAB, yielding BMD Z-scores similar to pretreatment levels. In AMAB, BMD Z-scores did catch up with pretreatment levels after long-term GAH treatment at the total hip and femoral neck, but not at the lumbar spine. A possible reason for this might be sought in the relatively low dosage of estradiol in the first stage of GAH. This hypothesis was studied in Chapter 5. In this chapter the effect on BMD development of different estradiol dosages was studied in trans girls. The study consisted of three groups who were either treated with 2 mg estradiol, 6 mg estradiol, or 100-200 μg ethinyl estradiol (EE). Treatment with 6 mg estradiol or EE was given to limit growth in a selective cohort of trans girls. All individuals had received GnRHa for at least one year prior to estradiol treatment, and started treatment before turning 18 years of age. After two years of GAH, a greater increase in BMD Z-score was found in the groups treated with EE than in the group treated with 2 mg estradiol. Moreover, Z-scores caught up with pretreatment levels in the 6 mg and EE group at the lumbar spine, at the total hip, and at the femoral neck. In the 2 mg group catch up of BMD Z-score was only seen at the total hip. This suggests that BMD development is indeed associated with estradiol dosage. An alternative explanation for the persistently low lumbar spine BMD Z-score during GAH treatment is considered in Chapter 6. This study demonstrated an inverse relationship between BMD Z-score and age in AMAB diagnosed with gender dysphoria who had not yet started medical gender-affirming treatment. This relationship was not observed in AFAB. This finding suggests that other factors besides medical treatment affect BMD development in AMAB. We found that the decrease in BMD Z-score was partly attributable to a decrease in height-adjusted lean mass Z-score in AMAB, indicating physical exercise is a key factor in the downward sloping curve of BMD Z-score during treatment as well. This explanation was also suggested by others. Chapter 7 showed that sex-specific development of hip bone geometry during GAH treatment followed the course of the gender identity only when puberty suppression was started in early puberty. When puberty suppression was started longer after onset of puberty bone geometry followed the trajectory of the sex assigned at birth. This study deepens our understanding of bone strength in transgender individuals as well as sex-specific bone development during puberty in the general population. Bone geometry is an important part of bone strength, so this could hold implications for fracture risk in transgender individuals. As trans women starting puberty suppression in early puberty obtain bone geometry similar to cis women, fracture risk might be higher compared to trans women starting in late puberty. In a broader perspective, the results suggest there is a particular critical window in early puberty during which the main effects of sex hormones on bone geometry are exerted. Strengths & limitations Firstly, this thesis provides an overview of general characteristics, diagnostic and treatment trajectories in a relatively new, and growing, population. All data originate from the gender identity clinic of the Amsterdam UMC, location Vrije Universiteit Amsterdam, the largest pediatric gender identity clinic in the Netherlands. This center has provided transgender health care to the majority of transgender people medically transitioning. Thus, results from this center are a fair representation of transgender adolescents in the Netherlands. People were treated uniformly according to the well documented Dutch Protocol, and data were collected and stored systematically. Another valuable asset is the unmatched duration of follow-up in most studies. As the Dutch Protocol originates in this gender identity clinic there are 20 years’ worth of data for studying trends and long-term follow-up. Secondly, linking our dataset with data from Statistics Netherlands (CBS), the institution collecting nationwide statistical information about the inhabitants of the Netherlands, is a solid component of this thesis. Thirdly, it gained much needed insight into bone health in both transgender adolescents not yet receiving medical treatment, and in those treated with puberty suppression followed by GAH. The latter is of particular interest, as long-term effects of puberty suppression and subsequent GAH on bone health were unknown, yet highly anticipated. A limiting feature is the retrospective design of most studies, restricting research questions by missing data which were either collected incompletely, such as vitamin D concentrations, or not collected at all, such as physical exercise scores. Moreover, all conclusions were drawn from quantitative data. For example, continuation of treatment was assessed based on issued prescriptions, but whether discontinuation was related to regret was not investigated with qualitative interviews. Additionally, BMD Z-score is not the sole factor contributing to the perhaps more clinically relevant parameter of fracture risk which was not studied. Lastly, although studying a population from one center has its advantages as mentioned, this also limits the applicability to practice outside the Netherlands. Clinical implications of this thesis Chapter 2 shows that treatment trajectories, including the start of medical treatment, and desire for gender-affirming surgeries are diverse and subject to individual characteristics and legal requirements. Adolescent transgender health care should be tailored to fit individual needs. After the abrogation of a Dutch law, allowing people to change their legal sex without having to undergo gonadectomy, a much larger proportion of individuals decided to keep their reproductive system in situ. This has considerable consequences for preservation of fertility potential. Additionally, AFAB might have changed gynecological needs, such as suppression of menstruation, evaluation of abnormal uterine bleeding, or obstetric care. Furthermore, as long-term effects of GAH on the reproductive organs are largely unknown, clinical providers should make persons opting for this approach aware of population screening programs for breast and cervical cancer. The high percentage of continuation of GAH found in Chapter 3 is reassuring and suggests a low risk of regret in individuals starting treatment in adolescence. It should be noted that all participants had a diagnostic assessment prior to treatment initiation and were professionally supported during their transition. After comprehensive assessment by a mental health professional with ongoing involvement of a multidisciplinary team, risk of regret should not be a reason to withhold medical treatment. The most pressing issues regarding long-term bone health in transgender adolescents treated with puberty suppression are answered in Chapter 4. Results in AFAB are reassuring as BMD Z-scores had caught up with pretreatment levels after long-term GAH. Provided there are no other indications for DXA, routine DXA-scans in this population, that started medical treatment in late puberty, do not seem necessary. Yet, conclusions drawn on the lumbar spine region in AMAB are less encouraging. During treatment with GnRHa BMD Z-score decreased, without compensatory catch-up after long-term GAH. However, Chapter 5 and Chapter 6 provided possible explanations for this decrease, as BMD development was associated with a lower dosage of estradiol during GAH treatment and with lower lean mass already before the start of puberty suppression. This last finding may be a result of differences in lifestyle, more particular in levels of physical activity that promote bone accumulation, between cisgender boys and transgender girls. Decreased lean mass, however, did not explain the diminished Z-scores completely. Other factors, such as deficient sunlight exposure resulting in low vitamin D, or reduced calcium intake, might hamper proper bone mass development in these individuals. Altogether, in AMAB the lumbar spine region remains a point of concern. Lifestyle factors and estradiol treatment should be optimized, stimulating maximum BMD accrual to mitigate the risk of osteoporosis later in life. It is remarkable that an ever-growing group of healthy children and adolescents have reduced bone mass accrual during puberty, possibly due to altered lifestyle, but claiming the treatment with puberty suppression in itself alone results in poor bone health would be erroneous. It is of great importance that transgender adolescents at risk of inadequate bone mass development are recognized early on, and properly advised even before the start of medical treatment. Routine DXA scans remain important to monitor bone health in AMAB and should be done at least before start of GnRHa and GAH and until an adequate peak bone mass has been achieved. Societal considerations Although advocated by many, the Dutch Protocol has received heavy criticism as well. At the root of the criticism is a fear for adverse physical, cognitive, and psychosocial outcomes. Additionally, some raise ethical issues, challenging the treatment of minors for gender dysphoria on moral grounds, and questioning whether young adolescents are competent to make an informed decision on a life-changing treatment. In contrast, others have condemned the Dutch Protocol because of the crucial part the psychological assessment plays. They consider fulfilling certain diagnostic criteria required to start treatment as “gatekeeping”: a situation when healthcare providers control access to care by adhering to strict criteria. A different approach is the so-called “informed consent model”. This model puts emphasis on respecting people’s autonomy and decision-making capacity, and intends to reduce barriers, including waiting lists, to access medical transgender care. It does not require evaluation of mental health prior to starting medical transgender care. Although waiting lists to enter gender identity clinics constitute a major obstacle, a diagnostic assessment seems of added value as Chapter 2 showed that the majority of people not starting GnRHa did so because a diagnosis of gender dysphoria could not be established. It is unclear how many of these individuals would have started puberty suppression in an informed consent model, but possibly more might have started who would have decided to detransition later on. Although detransitioning in later life is by no means ruled out by a diagnostic and psychological evaluation prior to initiating treatment, Chapter 3 did find that very few individuals who started puberty suppression in adolescence followed by GAH subsequently stopped using GAH. This result can only be translated to individuals following the diagnostic approach according to the Dutch Protocol, including psychological assessment, but is reassuring regarding the risk of detransitioning in this particular population. An increase in referrals amongst children and adolescents, and a sharp rise in pubertal AFAB requesting care as noted in Chapter 2 adds another degree of uncertainty to treating transgender adolescents: whether or not recently referred adolescents are comparable to the ones for whom the Dutch Protocol was originally designed and evaluated. Although a satisfactory explanation for the ratio strongly favoring AFAB has not yet been found, we can speculate about possible reasons. First, gender dysphoria might occur more frequently in AFAB. Reasons for this may include biological factors such as in utero exposure to abnormal levels of sex steroids or endocrine-disruptive chemicals, but studies have not corroborated these thoughts. Second, the observed imbalance could reflect sociocultural precepts either urging AFAB to seek care more often or at an earlier age, and/or restricting AMAB to seek care. Possibly, AFAB can find their way to gender identity clinics more easily because in most Western societies it is more socially accepted for AFAB to present as gender diverse than for AMAB. A patriarchal society in which masculinity is valued above femininity, or in which being male is the path of least resistance, might play a role in this. However, it is unclear how this relates to the ratio originally favoring AMAB just after the introduction of the Dutch Protocol. Alternatively, the role of (social) media should be considered. Transgender and gender diverse-related items are appearing in the media exponentially, and young people are increasingly exposed to (social) media. Perhaps young AFAB are more susceptible to this coverage, and triggered to question their own gender identity, but reliable evidence on this is lacking completely. In Chapter 2 it was shown that the proportions of adolescent AFAB starting GAH were more or less similar for individuals first visiting before or after age 10. Additionally, Chapter 3 demonstrated continuation of GAH was similar in AFAB starting GAH before 2012 and after 2012. This suggests there are no major differences in persistence of gender dysphoria after start of puberty suppression in prepubertal and pubertal AFAB, nor after start of GAH in more recently referred AFAB. Future studies Although this thesis greatly contributes to the knowledge on long-term outcomes of the Dutch Protocol, long-term studies on gender-affirming treatment in adolescents, as well as the capacity of adolescents to make informed decisions about their healthcare, remain crucial to improve safety of transgender health care to adolescents. Unexplored areas of interest include effects on cardiovascular health and cognitive functioning. Another objective should be the long-term efficacy of the Dutch Protocol, investigating whether starting medical treatment in adolescence improves mental health outcomes and well-being. As the number of referrals keeps increasing, it is important to find cogent reasons for the shift in ratio of referrals to, mainly pubertal, AFAB, and whether there are different characteristics of these individuals that challenge diagnostic certainty and may influence long-term outcomes of the current protocol. Furthermore, relatively more people choose to not undergo gonadectomy and keep their reproductive organs in situ. Long-term outcomes regarding the potential risk of malignant transformation due to continued exposure to exogenous sex hormones need to be studied. Further opportunities for exploration are the persistent low lumbar spine BMD Z-score in AMAB after long-term GAH, and potential consequences in terms of fracture risk. Conclusion As a whole, this thesis shows a low risk of discontinuing GAH in individuals starting puberty suppression in adolescence prior to GAH and largely refutes the historical concerns regarding long-term bone health. It adds scientific arguments to a discussion now mainly depending on expert opinion, political and ideological concerns. This thesis could help guide the highly polarized public and political debate regarding treatment for gender dysphoria in adolescents. However, more long-term studies, in all domains, are required to inform medical practice, policy decisions, and public discourse, ultimately contributing to a more just and evidence-based treatment perspective for transgender and gender diverse adolescents.