The Cass Review Into Gender Identity Services For Children - Part 7

Bad science and the alternative to drugs.

I searched psychology on Unsplash and it gave me this wonderful photo of eggs. Photo by Tengyart on Unsplash

This is part 6 of my series looking at the Cass review into gender identity services for children in the UK. You can find the other parts here (I will update as I add sections):

- Introduction

- Part One

- Part Two

- Part Three

- Part Four

- Part Five

- Part Six

- Part Seven

When you talk policy, it’s important to look at both the words that people write and what those words actually mean in practice. The words contained in a document are interesting to understand the author’s thoughts and beliefs, but when we’re talking about advice to governments on what they should do to people’s lives, the most important thing is what the practical outcomes of those recommendations might mean. If someone tells the government to build a new highway through the middle of a city, but doesn’t mention all of the homes that are going to be demolished to achieve this, it’s a very bad sign.

The Cass review is atrocious at doing this. Many of the recommendations are positioned as if they are context-free, with no further impact than the simple words that the authors have written. Take the recommendations for treatment. If you only read the specific words, you might forgiven for thinking that there is not treatment that the authors of the Cass review believe should be given to children specifically for gender dysphoria (as opposed to, say, a diagnosis of depression). However, if you look carefully, the review actually does implicitly recommend that children with gender dysphoria should receive one for of care: completely unproven psychological/psychosocial therapies.

This comes from a fairly simple fact - there are currently two main ways for transgender children to access care for their gender dysphoria. Firstly, medical. But the Cass review has recommended against all medical interventions in children, going so far as to advocate for a ban on some treatments. Indeed, it is now a criminal offence for a trans child in the UK to possess puberty blockers (there is no prohibition for cisgender children). Despite the review itself finding that psychological/psychosocial interventions are lacking any evidence whatsoever, they are the de facto recommendation of the document.

Let’s look at one of the key hidden recommendations of the Cass review, and why it is extremely dubious from a scientific perspective.

Recommending Psychosocial Therapies

The first point that’s important is to show that this is indeed a recommendation of the review. There are basically three treatment modalities that the Cass review discusses:

1. Medical management of gender dysphoria, which primarily involves some combination of puberty blockers and gender-affirming hormones after a period of assessment which usually includes psychological and psychiatric consultations.

2. Psychological and psychosocial interventions for gender dysphoria, which are a diverse range of possible therapies aimed at improving dysphoria through counselling and similar, without necessarily encouraging or even allowing transition to take place.

3. Psychiatric and psychological care for conditions other than gender dysphoria (depression, anxiety, etc). The review argues that this aspect of care is majorly deficient in the UK, which as I’ve discussed is very confusing given that their own data contradicts this belief.

In other words, a child going to their doctor or mental health professional will have three options in front of them in the UK. I’m not counting social transition, because it is not really under the control of medical professionals.

The Cass review recommends largely if not entirely banning the first of these treatment modalities. That is a fairly uncontroversial fact - the only person who seems to disagree is whoever wrote the FAQ page for the review. Psychiatric and psychological care for issues other than gender dysphoria are already used broadly for trans kids - again, this is not controversial, and is shown by the Cass review’s own data. Thus, the only specific treatment that the review recommends is that children see a psychologist or similar mental health professional to treat their gender dysphoria:

“Beyond this first line approach, it is important to understand how specific therapeutic modalities may help the core gender dysphoria and bodily distress. One of the given rationales for puberty blockers is that they may improve gender dysphoria or overall mental health. The evidence to date does not provide strong support for this (see Chapter 14). Furthermore, even after masculinising/feminising hormones, dysphoria may still persist. Therefore, it is important to explore other approaches for addressing the gender-related distress, which in itself is debilitating. These may be of value regardless of whether or not an endocrine pathway is chosen.” (Cass Review, Page 155, Chapter 11: Psychological and Psychosocial interventions)

So, the Cass review recommended psychological/social interventions as the most reasonable treatment available to children to address gender dysphoria. Let’s look at the data that the authors based that decision on.

The Science

The primary source for data on these psychological/social interventions for gender dysphoria is, as with most of the Cass review, a University of York systematic review on the topic.

On the whole, these reviews have been fairly good. However, the review into psychological/social interventions had some bizarre and unexplained differences to the other systematic reviews included in the Cass report which make it much less reliable.

One big issue is that all of the other systematic reviews excluded low-quality evidence. This is, as I discussed previously, a defensible choice. The main problem with low-quality data is that it doesn’t really add to your conclusions, and so discarding it isn’t a particularly problematic thing to do.

However, the systematic review into psychological/social interventions did not exclude low-quality studies. In fact, of the 10 studies identified, the University of York authors considered 9 to be low-quality - all of these were included as direct evidence both in the systematic review and the main Cass review document. This massive protocol deviation is not explained nor referred to, and severely undermines the Cass review’s discussion of these interventions.

The York systematic review also used a different scoring system to rate quality in this one review than all the others. In most of their papers, they used the Newcastle-Ottawa scale, but for this one study they instead used the Mixed-Methods Appraisal Tool. This likely changed their findings quite a bit, but is not explained or discussed in the paper. As I’ve noted before, it’s not a problem to use one methodology or the other, but if you’re going to pick and choose which scoring system to use you have to at least explain why you’ve made that choice.

So this is the worst systematic review of the bunch by far. That aside, what were the studies the York team found? There were only 10 of them:

1. This case study of attachment-based therapy on a single transgender teen. Astonishingly, this was the study considered to be the highest quality of the research the team reviewed.

2. This review of a waitlist-based intervention where trans youth were triaged when they were put onto the waitlist to see a doctor about gender dysphoria rather than just going on the waitlist as usual. The paper found that being seen more quickly had some benefits for mental health in these teens.

3. This small before/after study of 8 trans teens which looked at whether having a psychologist affirm their trans identity was helpful in improving their depression.

4. This study of online mindfulness training. Of the 41 teens who started the study, 26 finished it. Their scores improved somewhat on a range of self-reported questionnaires for depression and anxiety.

5. This paper which was also included in the review on puberty blockers. It looked at a fairly large sample of trans teens who either got psychological support or psychological support plus blockers. The average rating on a 100-point scale at the start of the study for both groups was 58 - in teens given psychological support, this improved to 61, but improved substantially more for kids who got blockers as well to a score of 67.

6. Another tiny before/after study of 11 children who attended group peer-support sessions with other gender diverse kids in the UK. There were minor improvements to self-reported mental health seen in this paper.

7. A small evaluation of online cognitive behavioral therapy in 207 transgender teens with a large control group of cisgender teens. Of these, only 14 trans teens finished the follow-up surveys. The study found that online CBT had benefits for cisgender children, but that it had a small negative impact on the depression scores of trans kids.

8. A moderately sized study looking at 200 children, of whom 35 were trans, who were admitted to psychiatric hospitals for reasons of mental health. All of the children saw improvements in their mental health from the time they were admitted to when they were discharged.

9. This before/after study of an online self-affirmation program for LGBTQ+ teens and straight allies. Of this group, 21 people identified as trans or gender fluid. This subgroup saw small improvements in their self-rated confidence and self-acceptance after the program.

10. This post-hoc analysis of data from randomized trials on psychotherapy. The authors pooled data from 4 RCTs, and identified 64 trans or gender diverse kids in the sample. These children did see some improvements as part of the trials, but substantially less so than their cisgender peers.

These findings are, in a word, mixed. Most of these papers are undeniably low-quality, and as with the low-quality literature for other aspects of healthcare for trans teens don’t really add much to the literature. In addition, the results were pretty contradictory - while some of these psychological and psychosocial interventions found that trans teens reported better mental health after the intervention, some of the studies showed the opposite effect. One trial of traditional psychological care for trans youth - the study of online CBT - found that it might make depression worse.

In addition, this review found no data whatsoever looking at interventions aimed at improving gender dysphoria - the entire point of the document. The majority of the studies looked at traditional psychological therapy but in a subgroup of transgender children. We already know that CBT is useful for depression. All this review shows is that traditional psychological therapies may not be effective for transgender teens, which also casts doubt on one of the speculative arguments of the Cass review that bad mental health turns children trans.

All in all, the data doesn’t give us much to go on. It’s possible that some of these interventions might help - for example, peer support groups for trans teens sound like a reasonable thing to do - but there’s no strong evidence that they have benefits. There’s also direct evidence from paper 5 on that list that psychological therapy alone is substantially less beneficial for trans teens than psychological therapy in addition to medications.

At best, we could say that psychological and psychosocial interventions could theoretically have benefits for trans teens, but that they seem less effective than medical assistance. In addition, the data is so weak that there’s very little you can reasonably say about them at all.

The Interpretation

What’s shocking here is not the scientific data. The studies are small, poorly-funded, and incomplete - that’s true of a lot of papers on trans healthcare, particularly in kids. The shocking part is the incredibly dubious - arguably even pseudoscientific - interpretation by the authors of the Cass review.

The protocol changes for this one systematic review are extremely worrying. If the Cass review authors had treated this in the same way that they did for all of the other systematic reviews, they would have discarded 9/10 studies and been left with a single case study in one child to discuss. You can’t just pick and choose your methodology when it’s convenient, and there’s no reported reason that psychological therapies should be based on lower-quality evidence than anything else.

For a document that spends literally thousands of words lamenting the quality of evidence for trans healthcare, it is startling to see that the only therapies that Cass recommends are the ones with by far the worst evidence around.

Moreover, the interpretation of this data is starkly different to the studies on medical therapies. For medications, particularly hormones, the review argues that we need lengthy follow-up periods of years or even decades to see if they are harmful:

“A significant weakness of the studies evaluating psychological or psychosocial function [for trans teens given hormones] was the short follow-up interval, with many following-up for less than 1 year, and a smaller number for up to 3 years” (Cass review, page 184, square brackets added for clarification)

The single longest study of any psychological therapy that the review could find measured outcomes up to 6 months. Here’s what the review notes about this follow-up period:

“The studies focusing on psychological changes and/or psychosocial changes found improvements in a range of aspects such as resilience, self-compassion and self acceptance, as well as quality of life, global functioning, participation and well-being. Where there was adequate follow-up, studies found that many of these outcomes fell off over time. There was no indication across the studies of adverse or negative effects.” (Cass review, page 153)

The difference in interpretation here is quite remarkable. Follow-up of 3 to 6 months, for non-medical interventions, is considered “adequate”, while even evidence with 3 years of data is inadequate when it shows benefits associated with hormones.

This is one of the most scientifically problematic parts of the review, and it really undermines the entire rationale of the document. Medical interventions are assumed to have lifelong harms based on Reddit posts and studies on rats whose ovaries were removed. Meanwhile, any data for benefits has to be totally unassailable to be counted as good evidence.

But for non-medical interventions, particularly those that don’t support trans kids in their transition, literally any data will do. The only therapies recommended by the Cass review are supported by the lowest-quality evidence in the entire document.

And look, I’m not saying that none of these psychological therapies could be helpful. Some of them seem perfectly reasonable. Self-affirmation therapy and online peer-support sounds like a great way to support trans teens in their mental health, although you could’ve just gone to Tumblr to see that in action without the scientific papers. Interventions to reduce waitlists are almost always beneficial in every context - no one wants to be on a waitlist!

The problem is that the Cass review has treated evidence that disagrees with its recommendations completely differently to the data that supports them. It’s not unreasonable to discard low-quality studies, but you have to be consistent if you do so. You can’t accept any old garbage that agrees with you if you’ve already thrown away dozens of papers that don’t.

There are probably psychological and psychosocial interventions that can help trans kids, but by positioning these as the only option in the UK the Cass review has veered substantially away from the evidence. Which brings us to my final piece on the review: the consequences.