---
tags: "puberty blockers"
title: Fact-checking Michael Laidlaw on puberty blockers and bone density
description: Fact-checking Michael Laidlaw on puberty blockers and bone density
image: https://upload.wikimedia.org/wikipedia/commons/thumb/1/13/Steroidogenesis.svg/250px-Steroidogenesis.svg.png
---
# Fact-checking Michael Laidlaw on puberty blockers and bone density
Recently, a [Twitter thread authored by Michael Laidlaw](https://archive.is/Xyh6u) has been making the rounds. It purports to analyze the results of the [Tavistock study](https://www.medrxiv.org/content/10.1101/2020.12.01.20241653v1.full) with respect to bone density. Unfortunately, what he writes is highly deceptive.
In fact, he's gilding the lily a bit too much. If one were to take him at his word, we'd be inflicting severe disabilities on trans youth and one would have to start wondering how an international conspiracy across multiple countries and political systems underwritten by all major medical organizations could possibly happen. This alone should raise questions about the credibility of the arguments he puts forth.
Note that the following is not an evaluation of puberty blockers as a whole. There are medical guidelines much longer than this write-up that provide a more comprehensive picture and consider harms and benefits. I am simply writing this to explain how Laidlaw is, as they say, economical with the truth. While one can never be sure if falsehoods have their origins in ignorance or malice, this level of ignorance would be troubling in a medical professional, too. Regardless, as Laidlaw is habitually making statements that are at odds with the medical consensus on the treatment of trans youth, this should help readers weigh the reliability of his contributions.
# Z-scores and what they actually mean
The first thing he notes is that Z-scores see a drop over the duration of 2-3 years that looks massive and scary. But mostly, he just exploits that most laypeople won't know what a Z-score is, just something that seems to be related to bone density. Let us explore in detail what they actually are about.
It is well-known that bone mineral density (BMD) is accrued at a slower rate in a prepubertal state than after the onset of puberty, as sex steroids play a critical role in bone mineralization. This holds for an artificially delayed puberty also. Slower BMD accrual on puberty blockers is *expected*. What we're looking at is whether BMD catches up after going off blockers, something that Laidlaw does not explore.
A Z-score is the number of standard deviations (or fractions thereof) that BMD is above (if positive) or below (if negative) the average BMD of a reference population of the same chronological age. However, using chronological rather than developmental age does not result in a very useful metric for adolescents. In particular, if puberty is delayed (normally or artificially), then the Z-score of prepubertal teenagers will drop, as they accrue BMD at the slower prepubertal rate than a teenager who is going through puberty.
Note that while Z-scores are dropping, BMD is not. In fact, BMD shows a small increase relative to the baseline of that population. The drop in Z-scores indicates a reduced rate of increase, not a decrease in BMD.
However, once puberty kicks in for those with a later puberty, we expect that the gap closes again. Overall, studies generally show (with some caveats) that after going off puberty blockers and going through puberty, BMD starts catching up. This is the actually interesting part that clinicians are interested in; BMD stagnating during puberty suppression (or really, normally delayed puberty) is old news.
In short, a drop in Z-scores is exactly the expected outcome for a delayed puberty and nothing to be alarmed about, as long as BMD values themselves do not drop – which they didn't. The result that we would actually be interested in is the catch-up data, but the Tavistock study does not provide that, as many of the patients had only just started cross-sex hormones.
We can observe this effect in [this unrelated study of cis girls who were *not* on blockers](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4457937/), but simply started puberty at different ages (early vs. late onset was a little over two years apart). See [Figure 1](https://www.ncbi.nlm.nih.gov/corecgi/tileshop/tileshop.fcgi?p=PMC3&id=292258&s=52&r=1&c=1) in particular. Late onset was associated with about a 1.3 point drop in the Z-score on average, which is actually pretty close to the value in the Tavistock study after 2-3 years.
Another detail that Laidlaw misses out on is that BMD at *baseline* (i.e. before puberty suppression was initiated) was already lower than for cis controls. This is in line with [another recent study](https://academic.oup.com/jes/article/4/9/bvaa065/5866143) where BMD at baseline was also lower than in cis controls, and at least part of that was explained by the lack of physical activity that the patients had engaged in due to gender dysphoria. Thus, non-treatment may also be harmful for gender dysphoric adolescents, but because it discourages physical activity (another major factor affecting bone health during puberty). This is also something that needs to be taken into account when evaluating the medical ethics of puberty suppression.
# Quantifying BMD development
Laidlaw then goes on to show a qualitative curve that exaggerates bone density growth by about 1-2 orders of magnitude.
The line is near vertical during adolescence, which just isn't factually accurate.
In order to provide an accurate picture, see the age/BMD charts in [this paper](https://www.semanticscholar.org/paper/Bone-mineral-density-according-to-age%2C-bone-age%2C-in-Lim-Hwang/bde63212b8161da40fdb19bd5088834f81d651e9) about Korean school children aged 5-20. I'm using it simply because it is openly accessible on the internet, but the points I'm making are not specific to Korean youth and can be observed in other populations.
For your convenience, I've charted the values for the BMD at the lumbar spine for girls from that paper:
```vega
{
"$schema": "https://vega.github.io/schema/vega-lite/v4.json",
"description": "BMD (LS) at given age",
"data": {
"values": [
{"age": 5, "bmd": 0.644},
{"age": 6, "bmd": 0.665},
{"age": 7, "bmd": 0.664},
{"age": 8, "bmd": 0.733},
{"age": 9, "bmd": 0.779},
{"age": 10, "bmd": 0.813},
{"age": 11, "bmd": 0.871},
{"age": 12, "bmd": 0.964},
{"age": 13, "bmd": 1.038},
{"age": 14, "bmd": 1.095},
{"age": 15, "bmd": 1.124},
{"age": 16, "bmd": 1.117},
{"age": 17, "bmd": 1.155},
{"age": 18, "bmd": 1.154},
{"age": 19, "bmd": 1.183}
]
},
"mark": {
"type": "line",
"point": true
},
"encoding": {
"x": {
"field": "age",
"type": "ordinal",
"axis": {
"title": "Age"
}
},
"y": {
"field": "bmd",
"type": "quantitative",
"axis": {
"title": "BMD (LS)"
},
"scale": {"domain":[0, 1.4]}
}
}
}
```
As you can see, this is is a much flatter curve than what Laidlaw tries to sell us.
It remains flat from age 5-7, starts increasing at age 8 when the first girls enter puberty (and, importantly, adrenarche happens) and then goes up until about age 15, when its slope goes back down. Age 8-15 are the typical pubertal years for girls.
In fact, we can go and see what happens if we hypothetically were to flatten the curve after age 12 for 36 months and shift the rest to the right. I'm not saying that this happens, but when Laidlaw inserts his horizontal lines in the graph, this is what it would actually look like. This is just the corrected form of his graph.
```vega
{
"$schema": "https://vega.github.io/schema/vega-lite/v4.json",
"description": "BMD (LS) at given age",
"data": {
"values": [
{"series": "shifted", "age": 5, "bmd": 0.644},
{"series": "shifted", "age": 6, "bmd": 0.665},
{"series": "shifted", "age": 7, "bmd": 0.664},
{"series": "shifted", "age": 8, "bmd": 0.733},
{"series": "shifted", "age": 9, "bmd": 0.779},
{"series": "shifted", "age": 10, "bmd": 0.813},
{"series": "shifted", "age": 11, "bmd": 0.871},
{"series": "shifted", "age": 12, "bmd": 0.964},
{"series": "shifted", "age": 13, "bmd": 0.964},
{"series": "shifted", "age": 14, "bmd": 0.964},
{"series": "shifted", "age": 15, "bmd": 0.964},
{"series": "shifted", "age": 16, "bmd": 1.038},
{"series": "shifted", "age": 17, "bmd": 1.095},
{"series": "shifted", "age": 18, "bmd": 1.124},
{"series": "shifted", "age": 19, "bmd": 1.117},
{"series": "shifted", "age": 20, "bmd": 1.155},
{"series": "shifted", "age": 21, "bmd": 1.154},
{"series": "shifted", "age": 22, "bmd": 1.183},
{"series": "original", "age": 5, "bmd": 0.644},
{"series": "original", "age": 6, "bmd": 0.665},
{"series": "original", "age": 7, "bmd": 0.664},
{"series": "original", "age": 8, "bmd": 0.733},
{"series": "original", "age": 9, "bmd": 0.779},
{"series": "original", "age": 10, "bmd": 0.813},
{"series": "original", "age": 11, "bmd": 0.871},
{"series": "original", "age": 12, "bmd": 0.964},
{"series": "original", "age": 13, "bmd": 1.038},
{"series": "original", "age": 14, "bmd": 1.095},
{"series": "original", "age": 15, "bmd": 1.124},
{"series": "original", "age": 16, "bmd": 1.117},
{"series": "original", "age": 17, "bmd": 1.155},
{"series": "original", "age": 18, "bmd": 1.154},
{"series": "original", "age": 19, "bmd": 1.183}
]
},
"mark": {
"type": "line",
"point": true
},
"encoding": {
"x": {
"field": "age",
"type": "ordinal",
"axis": {
"title": "Age"
}
},
"y": {
"field": "bmd",
"type": "quantitative",
"axis": {
"title": "BMD (LS)"
},
"scale": {"domain":[0, 1.4]}
},
"color": {
"field" : "series",
"legend": {
"title": "Series"
},
"scale": {
"scheme": "category20"
}
}
}
}
```
You can see Laidlaw's diagram here. Compare the orange curve to the actual data from actual girls above, including the hypothetical flattening.
For the full story – which Laidlaw elides and which is the actually interesting part – I do need to point out that things still do depend on the amount of catch-up you get to do. For trans youth in particular, this depends on the type of hormone therapy they are getting. There are also arguments that unconditionally putting trans youth on puberty blockers until age 16 instead of starting cross-sex hormones earlier if the diagnosis is certain to be positive is difficult to justify from a point of medical ethics.
Note that we cannot guarantee that trans youth *will* reach the same BMD as their cis counterparts would have without. But this revolves around the catch-up question; that BMD will stagnate during pubertal suppression and see at most modest increases is a given.
Instead, it seems that the standard protocols for puberty suppression, which require not only monitoring bone health and identifying and counteracting problems (e.g. vitamin D deficiency and lack of physical activity) through medical, therapeutic, and social measures seem to be the wisest way to go for with trans youth.
# How does puberty suppression actually work?
I'll also note that Laidlaw is talking about "poisoning the pituitary". I have no idea what he means by that, because it has no relation to how puberty blockers work. While GnRH analogues work on the pituitary, they do not "poison" it anymore than hormonal contracepives do (which also act on the pituitary).
In adolescents and adults, the hypothalamus produces so-called gonadotropin-releasing hormone (GnRH), which signals the pituitary gland to make follicle stimulating hormone (FSH) and luteinizing hormone (LH). FSH/LH then either induce the menstrual cycle in ovaries; or in testes, FSH triggers spermatogenesis and LH causes the production of testosterone. This part of the endocrine system is called the hypothalamus-pituitary-gonadal axis, or HPGA for short. There are some further details on how the HPGA works, but they are not needed for understanding puberty suppression.
This is also the mechanism through which puberty begins. Prior to the onset of puberty, the hypothalamus/pituitary don't make GnRH or FSH/LH, respectively, and so the gonads remain dormant.
How do puberty blockers interact with that mechanism? Puberty blockers are so-called GnRH analogues. They were developed shortly after the discovery of GnRH, but originally with a different intent, namely to emulate the function of GnRH (hence why they are called analogues). As it turns out, however, GnRH (and therefore, GnRH analogues also) works in a pulsatile fashion. If you administer GnRH analogues in a sustained fashion, you get an initial flare up of FSH/LH secretion, after which FSH/LH levels drop as long as you are on GnRH analogues (which is why they are administered using an implant or depot injections).
The underlying mechanism is that sustained use of GnRH analogues desensitizes the GnRH receptors in the pituitary. Once you go off them, the pituitary starts decoding GnRH pulses again. It's like a button that rings a bell once when you press it. If you press it and keep it pressed, other people can press it as much as they want, it won't ring the bell again. However, once you release it, it can be used to ring the bell again and again.
This mechanism is used for puberty suppression (usually in conjunction with a mechanism to counteract the effects of the initial flare up). The long and short of it is that GnRH analogues, when used for puberty suppression, stop FSH/LH secretion, upon which the gonads return to their prepubertal inactive state.
(GnRH analogues can also be administered in a pulsatile fashion to stimulate FSH/LH secretion; this has applications in IVF or to induce puberty.)
There is no poisoning of the pituitary here. Poisoning implies injury or harm. There is no injury or harm that is being done to the pituitary gland. In fact, GnRH analogues are so widely used precisely because they have no known significant *direct* side effects, only those resulting *indirectly* from stimulating or suppressing the HPGA. While the indirect side effects can be significant, especially in adulthood, stimulation or supression of gonadal function *is* what you want when you use them, so they can accomplish that goal with minimal additional risks.
# Risk/benefit analysis
Finally, it is not necessary that puberty blockers are completely risk-free, only that the risks and harms are properly balanced against the harms of gender dysphoria, especially the high psychiatric morbidity associated with gender dysphoria. Puberty suppression is preceded by a harm/benefit analysis, as noted in [this Tavistock paper](https://pubmed.ncbi.nlm.nih.gov/27431339/):
> In an attempt to balance the benefits and risks of puberty suppression, and in light of all the available information and knowledge, our opinion is that the enlightened decision would be to allow puberty suppression **when the adverse outcomes of a lack of or delayed intervention outweigh the adverse outcomes of early intervention in terms of long-term risks for the child**. In other words, if allowing puberty to progress seems likely to harm the child in terms of psychosocial and mental wellbeing, puberty should be suspended.
(Emphasis mine.)
Sign in with Wallet
Connect another wallet