28

This study and this article are being cited as proof that children should be allowed to transition. It argues that social transition (there's no mention of medical transition) makes gender nonconforming children happy less depressed. Is the sample group large enough? Do these positive results come from constant love bombing and attention, which isn't commonly given to non-trans children? Also, by trans, do they mean just gender nonconforming? This study also seems to disprove that children need any sort of medical transition to be happy normal, which is nice. I think it ultimately proves children should be allowed to express themselves in gender nonconforming ways, which is obvious.

Edit: thanks to everyone for taking the time to chime in. Let's consider this study debunked, I don't think it brings up any meaningful information.

[This study](https://www.jaacap.org/article/S0890-8567%2816%2931941-4/fulltext) and [this article](https://archive.thinkprogress.org/trans-kids-socially-transition-study-529f34c1bb3b/) are being cited as proof that children should be allowed to transition. It argues that social transition (there's no mention of medical transition) makes gender nonconforming children ~~happy~~ less depressed. Is the sample group large enough? Do these positive results come from constant love bombing and attention, which isn't commonly given to non-trans children? Also, by trans, do they mean just gender nonconforming? This study also seems to disprove that children need any sort of medical transition to be ~~happy~~ normal, which is nice. I think it ultimately proves children should be allowed to express themselves in gender nonconforming ways, which is obvious. Edit: thanks to everyone for taking the time to chime in. Let's consider this study debunked, I don't think it brings up any meaningful information.

12 comments

[–] worried19 32 points (+32|-0)

To me, it seems obvious that if you let a GNC kid be themselves, they'll be happier. I was allowed to pretty much "live as a boy" as a kid. If my parents had tried to force femininity on me, I would have been extremely depressed and quite likely suicidal.

The problem with the trans movement is that they insist on telling these children they're the opposite sex instead of just letting them dress and cut their hair the way they want. Clothes, toys, and hobbies should not be gendered in the first place. It's not a pathology to be a feminine little boy or a masculine little girl. It makes me sick and disgusted that GNC people are considered so abnormal it's now become acceptable for doctors and therapists to medicalize us out of existence.

[–] Liandra 23 points (+23|-0)

To me, it seems obvious that if you let a GNC kid be themselves, they'll be happier.

Or like...any kid. Let kid wear what they are comfy with (if weather appropriate), let kid find their own interests/hobbies, as long as it is not dangerous, let kid have agenda over their own life/body.

One time my stepmother tried to force me into a hobby I was not interested in. I was not happy. They took me out and I got to do music instead. Was happy. Agree that stuff should not be gendered. I am so sad we are not over that already!

[–] zephyrean 20 points (+20|-0) Edited

It argues that social transition (there's no mention of medical transition) makes gender nonconforming children happy less depressed.

SHOCKING DISCOVERY! Letting children do whatever they want makes them happy in the short term!

with 2 control groups: age- and gender-matched controls and siblings of transgender children.

These are not controls and this study is bunk. A control group for "does transing out make groomed children happier" would be age- and sex-matched groomed children who aren't allowed to trans out but are otherwise afforded the same liberties (including, for example, picking out a new name, so if you were Mildred Higginbottom, you could be Gloriana Firestar or Yukino Mireyu now).

It's like if they were testing a cure for cancer and the question they were asking was "is having cancer better or worse than not having cancer", while what they should be asking is "is ivermectin better than chemo".

I'd also argue that the equivalent of """therapy""" fees in the control group should be spent on nice shit for the kid. Because the question is not "are groomed children worse or better off than non-groomed children" - we know they're worse off! - it's "our daughter has been groomed and feels like shit, we have X dollars, what should we do with the money to make her better?".

Parents similarly reported that their transgender children experienced

A munchie parent satisfaction survey means fuck all.

Is the sample group large enough?

Dunno, it's paywalled, but there are no controls so it doesn't matter.

[–] pennygadget 13 points (+13|-0)

SHOCKING DISCOVERY! Letting children do whatever they want makes them happy in the short term!

I didn't fit in at school and it caused me to be depressed as a kid. I would have been a VERY happy child if my parents had allowed me to spend my youth playing video games and watching TV all day. But that would have been a very bad thing for me in the long term.

Sadly, I think they don't care about long term impacts because the kids who medically transition before puberty likely won't live to see middle age

[–] firebird 13 points (+13|-0)

Like someone else said, they basically don't have the sort of control groups that would allow them to state that socially transitioning these kids is better than not doing that, or trying to help them another way. From what I could tell they simply compared to children that aren't in the same boat at all.

Is the sample group large enough?

It's quite small, in part because not everyone in the study completed everything the were studying.

Also, by trans, do they mean just gender nonconforming?

They use the term gender nonconforming in their conclusion, not dysphoric, so my hopes aren't high here.

That's just for the study though, the article is just a pile of bullshit it seems.

The new study shows, in “striking contrast,” that allowing this social transition can be greatly beneficial to these young people’s mental health.

It showed no such thing, as that simply wasn't studied.

The study also debunked a myth that parents of trans kids may underreport how their children might be struggling. In this case, they actually reported having higher levels of anxiety about their children than the children actually reported for themselves.

How exactly was this debunked?

And then they start quoting Jack Turban who starts making claims that he basically admits aren't proven through his choice of words.

[–] loyal_WBW 7 points (+7|-0) Edited

Yes, let kids express themselves and follow their interests. But I think adults calling children ‘trans’ or ‘gay’ or any other adulthood thing is completely off base.

To put that “study” in perspective, I recommend reading INVENTING TRANSGENDER CHILDREN AND YOUNG PEOPLE, a book edited by Michele Moore and Heather Brunskill-Evans, with chapters by experienced clinicians and other professionals.

[–] KarousTattoos 5 points (+5|-0)

Some flaws have already been pointed out above, I'd just like to add a few of my thoughts.

The minimum criteria for inclusion were:

To be included as a transgender participant in the present study, children (1) needed to identify as the gender opposite their natal sex in everyday life, (2) must have socially transitioned by using the pronoun associated with their asserted gender in all contexts, and (3) were enrolled in the study between March 2015 and February 2016 (when the present measures were included).

There is no definition what they mean by identifying as the gender opposite their natal sex but their definition of a social transition suggests it's all about stereotypes:

A social transition is a nonmedical decision to allow a child to change his/her first name, pronouns, hairstyle, and clothing in order to live everyday life as one’s asserted gender.

There's various issues with all of this, but to begin with, we all know women who were like this as girls and who were not considered to be the opposite sex just because they expressed strong preferences for opposite-sex stereotypes. But we do know that enforcing compliance with sex stereotypes and sex role stereotypes is harmful to children.

Also, there's no information as to whether the children have gender dysphoria or not. So, we cannot compare this paper to studies done with children diagnosed with that condition (which typically show that a social transition alone does not alleviate their distress in the longterm).

There is also no data on the length of a social transition at time of being surveyed. This matters because a typical finding is that each step of the affirmative approach alleviates distress temporarily, necessitating a move onto the next step. So social transition helps, for a while, then puberty blockers. Puberty blockers help, for a while, then cross-sex hormones. Cross-sex hormones help, for a while, then surgeries. Surgeries help, for a while, then?

Moving on:

Control groups were only the siblings of these children and "age- and gender-matched controls".

It's unclear whether the latter means children who are all gender-conforming or a mix of ordinary kids, some of whom will conform and some won't or if it's children who simply don't meet the above criteria regarding pronoun use and identifying as a gender. No clue.

In my view, the second control group is also mismatched, or at the very least a third one is missing: children who meet the above criteria but who have not socially transitioned or children who are allowed to freely be gender-non-conforming without being seen as trans for that.

Furthermore, if their subject group consists of those at various stages of a medical transition, they cannot reliably claim that a social transition alone makes these kids happier.

That's what jumps out at me right now.

[–] Cats4Science 5 points (+5|-0) Edited

While this studies conclusion states that trans children report similar rates of depression as non-trans children, I think the authors are being willfully misleading. I only looked at the abstract, but from the data presented they missed a key finding - that across all measures trans children had higher anxiety rates than their peers. While two of the measures didn’t reach the .05 significance level, they fell under .1. The third was at .002. This should have been noted.

Additionally, while the sample size is strong enough to support findings, the study design does not actually tell us the key debate piece - whether or not mental health outcomes improve as a result of transition. To do that, the authors would need to begin the study prior to transition and compare mental health indicators longitudinally.

Edit: Found this rebuttal article describing in detail the misleading findings - https://journals.sagepub.com/doi/full/10.1177/0024363919884799

[–] ProxyMusic 5 points (+5|-0) Edited

The assessments of the children's mental health in this dubious study was all done by the children's parents, usually the parent most involved in transing them. Moreover, the principal method of assessment was for the parents once a week to fill out short forms containing statements such as, "My child cries a lot" and "My child likes to spend time by themselves" and checking off whether these statements applied always, frequently, half the time, sometimes or never .

So the arbiters of whether the kids were happy or not were the same people who transed these poor kids in the first place. Whose judgment generally and objectivity about their own kids I wouldn't trust for a second.

Also, IIRC the results show that even by the piss-poor methodology and rough measures used, the trans-identified kids actually were judged to have significantly more anxiety than the children they were compared to (and they were more anxious than they themselves had been prior to being socially transed). Moreover, their levels of depression and self-worth stayed the same, and were no worse than the kids they were compared to - all of whom come from a generation of kids with unprecedentedly sky-high rates of depression and self-esteem problems.

https://journals.sagepub.com/doi/full/10.1177/0024363919884799

[–] NotCis 4 points (+4|-0) Edited

OK, I did what I could while multitasking at work. For the sake of simplifying, I'll refer to "transgender children" even though the proper term in my eyes is "trans-identifying children."

Takeaway: This study has some pretty serious limitations. It found that transgender children report the same rates of depression and slightly increased rates of anxiety compared to their peers, but only anxiety was marginally higher in the analysis. Parents of transgender children reported that their kids had slightly higher anxiety compared to what the children report. This study does NOT (as it acknowledges) show any causal relationship between social transition and less anxiety/depression (meaning that TRAs should not point to it as "proof" of the "benefits of transition.") The study itself acknowledges this limitation in the discussion section. All that it does show is that in a limited, relatively small sample with some serious methodological issues, transgender children who socially transition remain pretty similar to non-rigorous control groups of siblings and peers when you assess depression and anxiety. They don't do better and they don't do worse.

Methodological issues I noted:

  • This is a longitudinal study, meaning that children were followed over time. Despite the study mentioning "control groups," this is not a randomized controlled trial (the gold standard for analysis) and the control groups don't make much sense. Comparing "socially transitioned" trans children to their siblings doesn't tell you anything - it would be more valuable to compare them to non-socially-transitioned trans children. The study does not do this.

  • The study recruitment was part of a larger national study, but recruitment was nowhere near random - it was word of mouth and through summer camps and groups catering to trans children. This is not a representative sample of U.S. children by ANY means. For example, this study notes that the cohort skews high-income. I suspect it's also skewed by clusters of trans friends.

  • In addition, the sample of "transgender children" included some kids on hormone blockers, some on cross-sex hormones, and some on neither. Luckily the study did assess differences between these groups, but none of the findings (for anxiety and depression) were significant. What this means, though, is that the study wasn't purely assessing "social transition" - a lot of the kids were also on hormones. Is it possible that girls taking testosterone experienced the known side effect of euphoria, which could bias the results? This wasn't assessed in the study.

  • The tools used to assess anxiety and depression look okay - they seem to be validated instruments. However, the "Global Self-Worth Subscale" looks suspect; "In this subscale, children were presented with a description of 2 kinds of children (e.g. “some kids like the kind of person they are BUT other kids wish they were different”) and were asked to select which kind of child they are most like. Once the children made a selection, they were asked whether this was “sort of true” or “really true.” Responses were recoded to a scale from 1 to 4, such that scores of 1 indicated the lowest self-worth and 4 indicated the highest." I would need to see more research on this instrument, but it looks like a pretty big leap to say that 6-year-olds have "low self-worth" based on this incredibly vague question about "kids wishing they were different."

  • The sample is small - just 63 transgender children. This isn't TINY, but especially when you consider the sampling limitations, it is nowhere near rigorous enough for drawing major conclusions.

I definitely recommend reading the discussion section if you can access the full text. It does a decent job of explaining the study's limitations (better than I can do while multitasking on a conference call...). Bottom line, it's too small and methodologically limited to point to as "proof" that trans children do better when they socially transition. It shows nothing of the kind. It only shows that they remain similar to poorly designed "control groups" of their siblings and peers.

[–] scriptcrone 4 points (+4|-0) Edited

From the abstract alone, it's difficult to tell. The fact that it is a small, non-randomized study, already puts it in the band of low quality evidence.

Points I would look at:

  • How were the patients selected? Was this specified in advance and decided by protocol? Who was excluded from the study, and how were their outcomes likely to differ.
  • Did everyone who went into the study contribute to the final analysis? This seems to be a common problem with trans studies, large numbers of missing participants. The problem with missingness is that they are very unlikely to be missing completely at random. Missing participants often have poorer outcomes than participants who stay in. Worst case, they're missing because they are dead.
  • Are the groups being compared really comparable? In a randomized controlled trial, randomization creates comparable groups by random assignment drawing from a common pool. The idea is to compare two groups that are identical except for the one thing, the intervention of interest. In the absence of randomization, it's very likely the groups aren't comparable. Without comparability, we don't know whether the result is driven by the intervention or the difference in characteristics. Matching is a way of creating comparability, but it is limited to measurable factors--and it depends on how well those factors are measured and adjusted for. For instance, socioeconomic status is a complex factor usually reduced to one term in an analysis: personal or family income. Is income going to capture the entire effect; probably not. Small numbers limit the number of factors that can be adjusted for.
  • Do they have the numbers to support their conclusion. Particularly if they're claiming to not see a difference. The more measurements, the more precise the averages (smaller standard error, narrower confidence intervals). The smaller the standard error, the smaller the differences between groups that can be detected. If they're using a scale where 2 is a meaningful difference, but they only have the precision to detect a difference of 4, then they will wrongly conclude that there is no meaningful difference.
  • What was the question anyway? Hypotheses should be properly set up. Were they looking for differences? If they were looking for a difference and didn't find one, then the best they can conclude is that they did not detect a difference, not that there is no difference.
  • How were the outcomes measured? It is surprisingly difficult to measure things that people believe they know, like mood. To get good results, a scale has to be separately validated. Does it actually measure depression (generally done by looking at correlation with other measures of depression)? Is it any good at telling the difference between severely depressed and less severely depressed people? Can it pick up a change in the severity of depression? And has the validation been done in the group of people who are being studied (adult scales shouldn't be used in children without that extra step). If the scale has been validated, it can't be modified without re-validating.
  • What was the distribution of measurements? Patient reported outcomes are very susceptible to floor and ceiling effects. They won't see a change within a group or a difference between groups, if most people are already clustering at the extremes of a scale. And standard statistical tests require that assumptions are met, like measurements being normally distributed. Patient reported outcome measurements frequently aren't.

Edit: And PS. Decision-making does not rely on one study. The highest level of evidence is the systematic review (of randomized controlled trials by preference), that looks at the totality of the evidence, and appraises individual and overall quality. Advocates of inappropriate interventions frequently exploit the effect of poor science education, which leaves people with a half-formed understanding of falsification derived from physics classes, and makes them susceptible to being convinced that one contradictory study disproves the others, simply because it contradicts the finding. Results from RCTs commonly contradict non-randomized studies, but that is because RCTs, if well done, avoid the biases of observational studies--they are a better design.

[–] NotCis 2 points (+2|-0)

When I have time today, I will take a look and report back!