A new study from Dr. Kaltiala in Finland is being used to claim that trans care is not lifesaving. Experts say this is false, and the study itself shows trans care likely saves lives.
This so much reminds me of climate denial junk science … those endless articles about how the “Mediaeval Warm Period” was warmer than today, and how Global Warming “paused” in 1998 (turns out it never did).
Same sort of sleight of hand and dubious methodology. Same use of a few “skeptic” scientists (all on the fossil fuel payroll) to sow confusion and doubt. Same media spin and gullibility about each dodgy paper that managed to find its way into a journal somewhere.
The other sleight of hand is the dismissal of all mental health concerns **unless** they result in suicide. That is, thankfully, quite rare, and rare events lead to weaker, noisier statistics. However suicide attempts are much more common, suicidal ideation more common still, and feeling downright miserable and suffering for years without treatment more common again.
What on earth are those criteria for gender dysphoria?!?! I’m trans, and gender dysphoria almost ended my life. After transitioning, I haven’t experienced dysphoria in years. But I don’t even come CLOSE to qualify for gender dysphoria under even the new criteria. Apparently because I’m a woman who enjoys sports, doesn’t have much bottom dysphoria, and didn’t play ”mommy” as a kid, I somehow DIDN’T experience life threatening gender dysphoria? Neat.
Interesting that all these studies that supposedly come to anti-trans conclusions rely on superseded scientific data. If they're going to rely on moot data, I have some data to recommend in their next "study":
-Spontaneous Generation (That living things come from non-living things)
-That heat is a substance called "Caloric"
-Astrology
-that there are only five elements on the periodic table of elements: Earth, Fire, Wind, Water, and Human (or Heart if you ever watched Captain Planet).
(I would suggest "flat earth" but there are idiots out there who actually believe in that superseded theory)
I mean, if these transmisic bigots are going to use outdated science, why not go all the way?
Big surprise that anti-trans research had to rely on faulty premises, outdated diagnoses, and flawed methodology... and they 𝘴𝘵𝘪𝘭𝘭 had to use pretzel logic to reach the conclusion they wanted, because 𝘦𝘷𝘦𝘯 𝘵𝘩𝘦𝘯 the data were against them. Looks to me like they were trying to pull a Catch-22 on their own subjects- if you sought help, then you couldn't have been suicidal, and if you didn't seek help, then you weren't suffering enough to need it. Thus, you only "count" if you ended up in the ground! I'm sure they'd love that.
The study itself is OK as far as it goes. The spin put on it by anti-trans activists is ridiculous and shows why they aren't taken seriously in the scientific community as a whole.
If you wish to look at the specific effect of trans care, the relevant comparison (GR+ vs. GR-) isn't even made in the study. So no conclusions can be made on it.
As far as the results go when visits to psychiatric care are included (highly correlated with suicidal ideation), all this shows is that GR patients with suicidal ideation don't have a greater risk of proceeding to an actual suicide than non-GR patients (it's not an additional risk factor). Interesting, but there is a still a higher rate of suicidal ideation and hence of suicide among GR patients, and that is obviously why we want to improve outcomes.
A favor: It would be a lot easier for me to understand the excerpted paragraph under "Paper Still Shows Trans Care Saves Lives" if someone could tell me what the terms HR, GR, and GR+ represent. Thanks.
In a society where people were never taught critical thinking or studied the errors of logic, people interpret reality though their beliefs. If it doesn't support their beliefs, then it is wrong. It's extremely hard to get around this, unless their beliefs cause such a violent conflict with reality that they are forced to face it.
Erin, Thank you for a much needed deeper look at the methodology/statistical analysis of this study. FWIW, I would call the impact of the GR+ (adjusted HR of 3.2) versus GR- (adjusted HR 0.8) a four fold difference.
Also, a quick thought on how to view "significance." As you imply, the adopted cutoff of 0.5 is arbitrary but it's also mischaracterized as the measure of "statistical significance."
While not necessarily widely described as such, this cut off is being used as a measure of "statistical stability." Also as you point out, results can fall short of this 0.5 cutoff and still be "clinically significant." Sorry if this feels like a bit of a rant but it always bothers me to know that this is what is taught and it's largely that it is portrayed in reporting result but it's so much clearer to think of statistical stability/replicability and clinical significance. Time now for me to get off of this hobby horse.
Leor Sapir from his Manhattan Institute bio: "He holds a Ph.D. in Political Science from Boston College ... Since joining MI, Sapir has become a widely recognized thought leader on topics related to pediatric gender medicine." Who the fuck is Leor Sapir, with his his bullshit Ph.D. in politics, to offer "incisive analysis" on the "capture of American medical groups on gender medicine"? Who is he to lecture medical professionals on these issues?
God, statistics make my eyes cross. I only read this because I know transitioning saved my life. I didn't even know that was a major issue until I finally started feeling better ... after 14 years of suicidal depression. These studies are insane and I shouldn't have to die to prove their points either way. SMFH
Folks who play with statistics must be very boring people ... or bored. That's all I know. Only if those statistics favor their lives are they ever utilized for the betterment of anybody.
I appreciate your writing about individuals who have had such a negative impact far beyond their abilities. That Finish abuser has shown up in too many anti-trans game plans, references for studies and think tank speeches.
I'm loath to use Dr. with Kaltiala. She obviously hasn't discovered that she's playing with accusations in the mirror. It's fascinating in the sense of discovering a new horrible disease that needs treatment.
The part if the article in that right-wing rag which irritates me greatly is the use of anecdotal presumptions. That in Finland she believed the majority of new clinic patients were FtM or non-binary who only had to wait until post puberty and discover their sexuality. That extreme nonsense is both conversion style and sample bias far beyond anything else I can recall.
Decades of reading abstracts and scientific journal articles for my benefit has led me to see much more than 38 years ago when I first worked as a library assistant making abstract references on cards.
> Adjusted HRs for suicide mortality were 3.2 (95% CI 1.0 to 10.2; *p=0.05*) and 0.8 (95% CI 0.2 to 4.0; p=0.8), respectively.
This is a huge red flag for "reverse p-hacking" (adjusting the analysis until a result you don't like becomes nonsignificant). Certain kinds of reverse p-hacking tend to produce a p-value at or just above the significance threshold, in this case 0.05.
(That study found "selective reporting of nonsignificant results" for the subset of the scientific literature it studied, in behavioral ecology, but no smoking gun that it was due to reverse p-hacking. I'm citing it for the explanation of what reverse p-hacking is and what the signs of it would be, rather than that conclusion.)
The whole study is also possibly an example of CHARKing (Constructing Hypothesis After Results are Known): it used a historical data set for which that would be particularly easy to do. If they hadn't been able to make this data fit the desired narrative about suicide in trans populations being due to psychiatric comorbidities (rather than lack of treatment), they'd have found some other historical data that fit, or just wouldn't have published it.
Thank you Erin.
This so much reminds me of climate denial junk science … those endless articles about how the “Mediaeval Warm Period” was warmer than today, and how Global Warming “paused” in 1998 (turns out it never did).
Same sort of sleight of hand and dubious methodology. Same use of a few “skeptic” scientists (all on the fossil fuel payroll) to sow confusion and doubt. Same media spin and gullibility about each dodgy paper that managed to find its way into a journal somewhere.
The other sleight of hand is the dismissal of all mental health concerns **unless** they result in suicide. That is, thankfully, quite rare, and rare events lead to weaker, noisier statistics. However suicide attempts are much more common, suicidal ideation more common still, and feeling downright miserable and suffering for years without treatment more common again.
What on earth are those criteria for gender dysphoria?!?! I’m trans, and gender dysphoria almost ended my life. After transitioning, I haven’t experienced dysphoria in years. But I don’t even come CLOSE to qualify for gender dysphoria under even the new criteria. Apparently because I’m a woman who enjoys sports, doesn’t have much bottom dysphoria, and didn’t play ”mommy” as a kid, I somehow DIDN’T experience life threatening gender dysphoria? Neat.
Interesting that all these studies that supposedly come to anti-trans conclusions rely on superseded scientific data. If they're going to rely on moot data, I have some data to recommend in their next "study":
-Spontaneous Generation (That living things come from non-living things)
-That heat is a substance called "Caloric"
-Astrology
-that there are only five elements on the periodic table of elements: Earth, Fire, Wind, Water, and Human (or Heart if you ever watched Captain Planet).
(I would suggest "flat earth" but there are idiots out there who actually believe in that superseded theory)
I mean, if these transmisic bigots are going to use outdated science, why not go all the way?
Big surprise that anti-trans research had to rely on faulty premises, outdated diagnoses, and flawed methodology... and they 𝘴𝘵𝘪𝘭𝘭 had to use pretzel logic to reach the conclusion they wanted, because 𝘦𝘷𝘦𝘯 𝘵𝘩𝘦𝘯 the data were against them. Looks to me like they were trying to pull a Catch-22 on their own subjects- if you sought help, then you couldn't have been suicidal, and if you didn't seek help, then you weren't suffering enough to need it. Thus, you only "count" if you ended up in the ground! I'm sure they'd love that.
𝘏𝘢𝘤𝘬𝘴.
This article is a very important refutation of the misinterpretations of the Finnish study.
One small editing note: you mention Southern Poverty but neglect to add the full name Law Center (SPLC.)
The study itself is OK as far as it goes. The spin put on it by anti-trans activists is ridiculous and shows why they aren't taken seriously in the scientific community as a whole.
If you wish to look at the specific effect of trans care, the relevant comparison (GR+ vs. GR-) isn't even made in the study. So no conclusions can be made on it.
As far as the results go when visits to psychiatric care are included (highly correlated with suicidal ideation), all this shows is that GR patients with suicidal ideation don't have a greater risk of proceeding to an actual suicide than non-GR patients (it's not an additional risk factor). Interesting, but there is a still a higher rate of suicidal ideation and hence of suicide among GR patients, and that is obviously why we want to improve outcomes.
A favor: It would be a lot easier for me to understand the excerpted paragraph under "Paper Still Shows Trans Care Saves Lives" if someone could tell me what the terms HR, GR, and GR+ represent. Thanks.
In a society where people were never taught critical thinking or studied the errors of logic, people interpret reality though their beliefs. If it doesn't support their beliefs, then it is wrong. It's extremely hard to get around this, unless their beliefs cause such a violent conflict with reality that they are forced to face it.
Erin, Thank you for a much needed deeper look at the methodology/statistical analysis of this study. FWIW, I would call the impact of the GR+ (adjusted HR of 3.2) versus GR- (adjusted HR 0.8) a four fold difference.
Also, a quick thought on how to view "significance." As you imply, the adopted cutoff of 0.5 is arbitrary but it's also mischaracterized as the measure of "statistical significance."
While not necessarily widely described as such, this cut off is being used as a measure of "statistical stability." Also as you point out, results can fall short of this 0.5 cutoff and still be "clinically significant." Sorry if this feels like a bit of a rant but it always bothers me to know that this is what is taught and it's largely that it is portrayed in reporting result but it's so much clearer to think of statistical stability/replicability and clinical significance. Time now for me to get off of this hobby horse.
Leor Sapir from his Manhattan Institute bio: "He holds a Ph.D. in Political Science from Boston College ... Since joining MI, Sapir has become a widely recognized thought leader on topics related to pediatric gender medicine." Who the fuck is Leor Sapir, with his his bullshit Ph.D. in politics, to offer "incisive analysis" on the "capture of American medical groups on gender medicine"? Who is he to lecture medical professionals on these issues?
God, statistics make my eyes cross. I only read this because I know transitioning saved my life. I didn't even know that was a major issue until I finally started feeling better ... after 14 years of suicidal depression. These studies are insane and I shouldn't have to die to prove their points either way. SMFH
Folks who play with statistics must be very boring people ... or bored. That's all I know. Only if those statistics favor their lives are they ever utilized for the betterment of anybody.
I appreciate your writing about individuals who have had such a negative impact far beyond their abilities. That Finish abuser has shown up in too many anti-trans game plans, references for studies and think tank speeches.
I'm loath to use Dr. with Kaltiala. She obviously hasn't discovered that she's playing with accusations in the mirror. It's fascinating in the sense of discovering a new horrible disease that needs treatment.
The part if the article in that right-wing rag which irritates me greatly is the use of anecdotal presumptions. That in Finland she believed the majority of new clinic patients were FtM or non-binary who only had to wait until post puberty and discover their sexuality. That extreme nonsense is both conversion style and sample bias far beyond anything else I can recall.
Decades of reading abstracts and scientific journal articles for my benefit has led me to see much more than 38 years ago when I first worked as a library assistant making abstract references on cards.
> Adjusted HRs for suicide mortality were 3.2 (95% CI 1.0 to 10.2; *p=0.05*) and 0.8 (95% CI 0.2 to 4.0; p=0.8), respectively.
This is a huge red flag for "reverse p-hacking" (adjusting the analysis until a result you don't like becomes nonsignificant). Certain kinds of reverse p-hacking tend to produce a p-value at or just above the significance threshold, in this case 0.05.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6364929/
(That study found "selective reporting of nonsignificant results" for the subset of the scientific literature it studied, in behavioral ecology, but no smoking gun that it was due to reverse p-hacking. I'm citing it for the explanation of what reverse p-hacking is and what the signs of it would be, rather than that conclusion.)
Adjusting for supposed confounders is a common reverse p-hacking technique. Here's another particularly egregious example to show how that's done: https://twitter.com/0xdoug/status/1643993363623084038
The whole study is also possibly an example of CHARKing (Constructing Hypothesis After Results are Known): it used a historical data set for which that would be particularly easy to do. If they hadn't been able to make this data fit the desired narrative about suicide in trans populations being due to psychiatric comorbidities (rather than lack of treatment), they'd have found some other historical data that fit, or just wouldn't have published it.
Similar study coming out of Hogwarts - junk science