> “[W]hen we adjusted our models for receipt of PB/GAH [that is, puberty blockers and gender-affirming hormones] we saw that depression and suicidality significantly worsened among youth who had not (yet) started PB/GAH at 3 and 6 months relative to baseline levels.” So the action here, as the researcher admits, comes not from the kids in the treatment group getting better, but rather from the kids in the comparison group getting worse.
This is a bit of a perspective argument. According to the small study that was conducted, treatment kept mental health the same, while it declined for those that went without treatment.
Many kids in the control group dropped out. This left the kids who did not improve in the control group
>The vast majority of kids who dropped out of this study didn’t access GAM. By the end of the study, just 6/63 of the remaining kids who provided mental health data, or 9.5% of the study participants, were in the no-GAM group. Overall, according to the researchers’ data, 12/69 (17.4%) of the kids who were treated left the study, while 28/35 (80%) of the kids who weren’t treated left it
If the study had continued to get details from all the kids we probably would not have seen an improvement compared to the control group.
I am assuming the kids who stopped having the issues would have dropped out at a higher rate. If a kid is having an issue their parents would want them to remain in to help find a treatment for the kid.
I am not aware of any data on why kids dropped out, but it does call into question the accuracy of the results. When almost everyone drops out of a study you should start over or make sure you get the reasons why they dropped out so you can figure out if they stopped having the issues.
> I am assuming the kids who stopped having the issues would have dropped out at a higher rate. If a kid is having an issue their parents would want them to remain in to help find a treatment for the kid.
This seems like a reasonable theory, but it seems just as likely to me that as things get worse people are more likely to decide this isn't helping and drop out. Which goes to the point of the rest of your comment, the dropout rate calls the whole study into question. Otherwise we are left with guesswork and assumptions.
> I am assuming the kids who stopped having the issues would have dropped out at a higher rate.
Interesting. I suppose that could happen. I guess the opposite might also be true if they dropped out because they didn’t want to wait on therapy any longer and once they started therapy would no longer be allowed to participate in the control group.
> When almost everyone drops out of a study you should start over or make sure you get the reasons why they dropped out so you can figure out if they stopped having the issues.
Agreed. I’m not sure why the author didn’t lead with this point. This alone seems like a sufficient critique, along with the small sample size.
With this kind of attrition rate, I am impressed that the researchers dared to write it up and got it published ... Then I found it was published on "JAMA Network Open", is it even a serious journal?
The actual concern raised, buried in the article (which is itself only a commentary on a different study):
"So the action here, as the researcher admits, comes not from the kids in the treatment group getting better, but rather from the kids in the comparison group getting worse."
Isn't that how control groups are supposed to work? Your treatment should produce better outcomes than the lack of treatment. This appears to correctly be the case here. The fact that these treatments alone don't resolve gender dysphoria, is expected, they are one part of a process. Stopping a rapid decline is one huge step in the right direction.
Well, you can't really tell if the other group got worse as a whole because it says later in the article that over 80% of the kids not on puberty blockers dropped out of the study so at the end there were only 6 kids left in the study who weren't on puberty blockers.
There isn't clearly a control group here, but a cohort of 104 kids at the start (of whom seven were already receiving PB/GAH) reduced to 65 (of whom 57 were now receiving PB/GAH) by the end, in twelve months. The supplementary materials explain this in good detail, but it doesn't make intuitive sense to me.
Only 6 of 63 in the no treatment group provided data at the end of the study compared to 57 of 69 of the treated provided data. Then there's a bunch of other points.
I mean, I feel like you just stopped reading to come to that conclusion.
A proper control group should be defined as the state-of-the-art alternative, not as "no treatment". The latter is a common ploy used to push substandard, "me too" treatments.
What's the "state of the art alternative" to address gender dysphoria in adolescents? Certainly the alternative to puberty blockers or hormones most often advocated is doing nothing.
The author makes some perfectly reasonable points about flaws in the study like the large and uneven number of dropouts from the study, likelihood of confounding factors in the nonrandom selection process and oddities in its construction to the extent I'm not sure the study itself can be used as firm evidence of anything, but his insistence that the treatment group not getting worse like the control group can't be evidence for it being beneficial isn't justified either.
Shouldn't the control group get placebo treatment? Simply by knowing they have been "treated" could already make them feel better. When the effect you are looking for is a psychological effect, it is even more serious a research design problem.
A concern I've always had with gender affirming treatments is that there's a strong belief that it is a medical condition, and not a psychiatric problem. I'm not a doctor or anything like that, but it seems to be very much a psychological issue, not a medical issue. Hormone treatment, surgeries, etc. are like medication to cover up the symptoms, but unless mental health is introduced to treat the psychological side, I don't think persons will find relief.
I'm not in mind that there isn't some subset of population that would not benefit from these treatments. What I question is how large it actually is and how big is effect of peer pressure, media trendiness and in general being teenager going through puberty have. Or just children being children. In some ways there might even be over emphasis of gender roles with momentary interest of children.
Now hard part is how to solve these issues without invasive actions and how to separate those who could benefit from such.
I will remind the gentle reader that the originator of the lobotomy was awarded a Nobel Prize for this.
From Wikipedia:
"The originator of the procedure, Portuguese neurologist António Egas Moniz, shared the Nobel Prize for Physiology or Medicine of 1949 for the "discovery of the therapeutic value of leucotomy in certain psychoses", although the awarding of the prize has been subject to controversy."
This article's title is disingenuous and is guilty of the same thing it accuses the study of.
The actual content of this article is the following (TLDR)
1) That the study didn't show a reduction in negative health outcomes, rather the control group showed an increase in negative outcomes.
2) The author of the article disagrees that this is equivalent to the stated claim, and offers a number of alternative explanations why you might see this outcome in the data
3) The author offers some critiques around lack of data transparency and the statistical rigor of the study.
In the end the title is a semantic nitpick around the difference between "improved absolute health" and "improved relative health outcomes" and not the smoking gun of bad faith actors it implies.
It’s not bigotry to discuss facts, and people who hide child abuse with the claim that it is bigotry are child abusers who ought to have their crimes published and prosecuted under the law.
This is a bit of a perspective argument. According to the small study that was conducted, treatment kept mental health the same, while it declined for those that went without treatment.
Pertinent data table: https://cdn.substack.com/image/fetch/w_1456,c_limit,f_webp,q...
So, it's more of a semantic argument. Did it really improve mental health?
Relative to where they were at the start of the trial, no.
Relative to the decline in the control group, yes.