When you take exogenous Test, your body converts some of it to Estrogen. In the high estrogen state it definitely can make you irritable, short tempered, etc (basically PMS). You need to be self aware enough to recognize that none of those feelings are real. If you aren't you could definitely act out in violent ways if you already have those tendencies. Easy fix though, just take an aromatase inhibitor.
People with naturally high Test? idk, can't tell you, I'm not one of them.
I've tested naturally high for T. I don't think I'm particularly violent. I do have ADHD, and I am impulsive, though not to a pathological degree.
We could argue that T is associated with impulsivity and risk taking. That seems well supported and more likely to explain observed effects. If so, would it predict increased aggressive action in a subgroup predisposed to aggression without making them "feel" more aggressive? That seems like a more likely mechanism. I also wonder how well people judge their aggressiveness. The guy hitting on the barista at the coffee shop probably just sees themselves as friendly. But their actions could be perceived as aggression by others.
It seems like +T -> impulsivity / sensation seeking is the better mechanism here. Did that hold up for you on TRT?
There is no literature on test with FTM transition and violence? Seems like the best natural experiment going from very low to normal male T if we slice by hormone replacement
I’ve heard it’s more accurate to call testosterone a status seeking hormone, as opposed to a violence hormone. It could be that violence is so often a way that men can obtain status, so that’s why it’s used so often.
The many things I thought about in your superb technical breakdown were, behavioral analysis and differences in male/female brain, impulse control, development in utero, the stress levels of the mother during pregnancy, and then I also remembered your pit bull writings in the past!
Is there a large amount of data you can find that shows any correlation between the bites and neutering? Based on what you found here, it’s probably not connected?!! Sorry if you already did that, I don’t remember lol
I had an English Bulldog (not neutered/raised from puppy) who was the kindest boy, and later, one who **was neutered** (adopted at age two) who unfortunately had major behavioral problems. I reckon the key differences in these two cases were primarily genetic and also the upbringing environment aspects. The latter dog was fearful easily and this typically indicates lack of exposure to various noises, people and situations in early puppy stages.
Could be multivariate, cutting across both biology/chemistry and the nurture aspects. Like for instance, how we deal with stress, how we learn to approach, communicate and solve questions. Learning how to have good outcomes to difficult conversations is a skill.
For instance, on the specific issue of violence between the sexes I've developed a nuanced take, whereby there's various contextual related issues at play, such as cultural norms (the stats of wives burnt alive in "kitchen accidents" in India for dowry related reasons are obscene), people snapping after years of nagging ("breaking point") and how our economic sphere breaks down dreams so that when reality shows up, men snap and vent frustration on the nearest person.
Correlations of T with psychological effects (especially sex drive) are nonlinear. Deprive a man of T and on average, his libido goes WAY down. (This happens most often in treatment of prostate cancer by Lupron.) But among men in the normal T range, there isn’t much of an association. If any.
I'm unclear on the point you're making about Androgen Insensitivity Syndrome? Do they not commit crime at far lower rates than normal males? My understanding was that chromosomal males with CAIS were behaviourally basically the same as women.
This one doesn't specify criminality as being among the traits measured, but most other sex-differentiated psychological traits seem to point to those with CAIS being basically the same as women. Curious on your thoughts
My point was that it's not the testosterone itself that's to blame. People with CAIS have male-typical levels of testosterone, it just doesn't do anything. This makes it similar to the race comparison due to the SHBG point.
It's meant as kind of a reductio, but it's an ambiguous observation on its own.
Did you look at the impacts on people's behavior or subjective self-experience when they start taking testosterone? I've seen many claims of people feeling much more aggressive and violent when taking testosterone.
Second, is it possible there's some minimum level of testosterone, and so it's really the incremental testosterone that explains things (which would help explain the relatively small male-to-female testosterone gap, but the much larger male-to-female violent crime gap)? I know this was a timed post, so you probably didn't have room to go into that.
When your main argument is that we can't find strong evidence of testosterone leading to aggression, but we also don't know why men and women are so different in terms of violent crime, there's a Chesterton's fence sort of principle that makes me skeptical of your conclusion. I also know there is significant literature correlating in-utero testosterone exposure with more stereotypically male behavior though I'm far from an expert there.
I think there are also theories of high testosterone exposure leading to developmental changes that result in a high but comparatively smaller gap in resting or free testosterone after development.
Was this the paragraph on 'roid rage? There seems to be a trend and it's even statistically significant so I think it's fair to say that testosterone increases aggression even if it may not be the primary driver of aggression.
Also, I believe that many people receive testosterone that is not a steroid but I haven't looked at that study's methodology.
A small thought and maybe totally off base, inspired by life on the farm.
Male cattle are castrated young (less than 90 days) but at one year steer and bulls are much closer together wrt to size than the steers are to heifers. After puberty, I think the difference is not as impressive between steer and cows, but the cows have way more body fat. The bulls of course are huge. I bet there is tons of evidence in the veterinary literature about the differences between intact males and, ahem, the minus 2 crew.
There is androgenizing effect that takes place during pregnancy. It seems to stop some time before birth, but it is enough to prime the brain. The phenomenon is known from mixed-sex calf twins where the female twin tends to develop into a freemartin, a frequently infertile chromosomally female calf. It is thought to be a combination of testosteron and anti-Müllerian Hormone (AMH).
Totally on board/familiar with the androgenizing effect in utero. (Did not know about the freemartin situation!) But I guess my focus here is how impressive post fetal growth is despite castration, and how much closer it is to typical male growth pre puberty. All seems like the growth and little boy, or little steer/male, antics are put into motion by something other than direct mediation of testosterone or already set in motion by birth.
Sorry, but the question of why men are more violent than women is completely different and unrelated to the question of why some men end up in jail, and you really only looked at evidence for the latter, as if it applied to your first question too, which it doesn't.
You should not be comparing testosterone levels between men to figure out why they're more violent than women, you should be comparing testosterone levels between men and women. Which men have about 15 to 45x as much as women, which would have been the much more obvious comparison in the chart you made regarding offender rates, and would seem to fully explain it. It would seem it is the jump from female ranges (typically under 70 ng/dL) to male ranges (265 to 900) that makes the difference, not the small differences in amounts once you're in male range.
As to comparisons between males rather than comparisons between sexes, I don't think criminal conviction is at all a good proxy for violence or aggression levels. In fact it's a very bad proxy. It's a much better proxy for having low intelligence and low social and actual capital.
There is lots of male aggression and violence that is not only not criminalized but celebrated. There are plenty of men smart enough not to get caught, or with the resources to stay out of jail if they do. Old men have more acquired capital to lose and young men don't, plus young men have more opportunities and access to sex with young women, so there's your reason for why they're more violent: they have more to win and less to lose.
I'm married to someone who has committed what might be described as extreme violence and has killed several people. Not only is he not a criminal, he is showered with awards and badges of honor for it. Bc he was a soldier. Plenty of men in MMA or other combat sports, actual combat, and plenty of other socially sanctioned activities that are not showing up in your numbers based on criminals who are mostly just dumb and lacking social resources.
The question of why men are more violent is not a mystery. Ask any evolutionary or behavioral biologists that focuses on animals. It's because the fundamental "purpose" if you will of males is just to fight other males for access to mating. Both sexes commit violence to protect and defend their territory, themselves, and (in the case of females mostly) their young. But when they're not doing those things or looking for food, pretty much the ONLY thing males do is fight with each other.
The question of why some men end up convicted for crimes and some do not is much more easily addressed by looking at how smart they are, what resources they have available to not be convicted, how entitled they've been taught to feel, what they have to lose, what they have to gain, and what their lethal and/or injurious capacity is. A guy who is likely to lose most fights probably won't get into them. A guy with a reputation and lots of assets to protect probably won't risk them. A young guy with no assets, no women, a high physical capacity for inflicting injury, and low IQ and self control...he probably will.
>You should be comparing testosterone levels between men and women. Which men have about 15 to 45x as much as women
He did that, noting the 17x higher levels in men.
>The question of why men are more violent is not a mystery. Ask any evolutionary or behavioral biologists that focuses on animals. It's because the fundamental "purpose" if you will of males is just to fight other males for access to mating.
That explains the reason why the trait evolved, not the neurochemical process by which it's imparted.
> That explains the reason why the trait evolved, not the neurochemical process by which it's imparted.
Indeed, a lot of people seem to be missing this. Yes, we can have sex differences in behavior without testosterone, and sex difference-related selection on adolescent and later behaviors can impact kids' behaviors. There are major sex differences in behavior in the mammal-expected direction among many invertebrate species, and they lack testosterone. It doesn't need to be invoked!
I would assume so, given the drastic impact it has on women's libido when they take tiny itty bitty nothing amounts that are about a tenth or less of what men supplementing with it take.
I just think you posed two questions at the beginning of the piece...and then answered different ones. None of this really looks at men as compared to women yet that is literally the opening line of the piece, so I assumed that's what it was about.
I believe you haven't cited “Sex hormone levels and body hair growth in !Kung San and Kavango men from Namibia,” by Eike-Meinrad Winkler & Kerrin Christiansen, American Journal of Physical Anthropology, Vol. 92, Issue 2 (Oct. 1993), pp. 155-164. Or a study showing that Sub-Saharan males have the highest testosterone levels of any race, A. Richard et al’s “Racial variation in sex steroid hormone concentration in black and white men: a meta‐analysis.” (2014) Andrology, 2(3), 428-435. I believe you also didn't cite See “Testosterone, Sex Hormone-Binding Globulin, and Body Composition in Young Adult African American and Caucasian Men,” by Stephen J. Winters et al, W.B. Saunders Company (2001). Forgive me if I did miss any of these studies and you did discuss them. I would be interested in knowing whether you think they are flawed in any specific sense: I've cited them all in my work and I certainly want to provide my own readers with a balanced assessment of their findings and what they imply.
As to why or whether men are more violent than women I don’t think that’s really the case. For example in unidirectional intimate partner violence women initiate it at about twice the rate as men, while in intimate partner homicide women commit it at roughly half the rate. People are conflating physical consequences with the decision to initiate it and it’s in the physical consequences that men have the lead.
The greater propensity for violence in men is clearly innate even from the most superficial life observations in childhood (before boys are more capable than girls at all) and adulthood, regardless of whether changes in testosterone can significantly alter aggression all by themselves. Without a background in all the relevant fields, I would simply guess that this arises from masculinized features of the brain that arise during sex differentiation.
How do these studies account for the testosterone-suppressive effects of opioids in the criminal population? Large proportion are using fentanyl plus other strong opiates and the men in prison know this, ask for T testing because they want to bulk up before a federal sentence. Older data would have heroin as opiate of choice. Violent behaviour persists with concomitant stimulant use and drug-induced psychosis, confounding T levels.
Hormones have both activational effects (at the time they are released) and organizational effects (that affect development and thus can persist after levels change). Sex hormones not only have activational effects but also organizational effects in utero, in infancy, and in puberty. Could you elaborate more on why you feel you ruled out prenatal effects? It looks like in your graph there was evidence for this via the digit ratios, yes? And it doesn't seem like you considered organizational effects during puberty or mini-puberty (the testosterone spike in infancy).
There is a sex difference in rough-and-tumble play in children as well as juvenile animals. In this period, there are not sex differences in androgens. However, androgens are implicated in this difference due to prenatal organizational effects. Girls with congenital adrenal hyperplasia have more rough-and-tumble play. Animal experiments have modified levels of rough-and-tumble play through manipulating androgen levels in utero. Yet if you tried to correlate rough-and-tumble play with testosterone at the time, you wouldn't find anything.
I'm not a statistician and I might have missed something critical in your post. If so, I apologize. But are we allowed to control for drug use if it's on the causal pathway?
+T -> increased reward seeing including drug use and sex -> violence ?
Hi! Great article.
Here is my personal experience with TRT:
Yes it does make you angry, but not exactly.
When you take exogenous Test, your body converts some of it to Estrogen. In the high estrogen state it definitely can make you irritable, short tempered, etc (basically PMS). You need to be self aware enough to recognize that none of those feelings are real. If you aren't you could definitely act out in violent ways if you already have those tendencies. Easy fix though, just take an aromatase inhibitor.
People with naturally high Test? idk, can't tell you, I'm not one of them.
I've tested naturally high for T. I don't think I'm particularly violent. I do have ADHD, and I am impulsive, though not to a pathological degree.
We could argue that T is associated with impulsivity and risk taking. That seems well supported and more likely to explain observed effects. If so, would it predict increased aggressive action in a subgroup predisposed to aggression without making them "feel" more aggressive? That seems like a more likely mechanism. I also wonder how well people judge their aggressiveness. The guy hitting on the barista at the coffee shop probably just sees themselves as friendly. But their actions could be perceived as aggression by others.
It seems like +T -> impulsivity / sensation seeking is the better mechanism here. Did that hold up for you on TRT?
There is no literature on test with FTM transition and violence? Seems like the best natural experiment going from very low to normal male T if we slice by hormone replacement
There's not enough literature yet. There is literature on sexuality and violence. I'll add something on that.
Very convincing! What do you think it is instead? Just genes directing brain structure development, or something else?
I'm really not sure! Whatever it is, it's something sex-related and developmental that isn't just testosterone-mediated.
I’ve heard it’s more accurate to call testosterone a status seeking hormone, as opposed to a violence hormone. It could be that violence is so often a way that men can obtain status, so that’s why it’s used so often.
The many things I thought about in your superb technical breakdown were, behavioral analysis and differences in male/female brain, impulse control, development in utero, the stress levels of the mother during pregnancy, and then I also remembered your pit bull writings in the past!
Is there a large amount of data you can find that shows any correlation between the bites and neutering? Based on what you found here, it’s probably not connected?!! Sorry if you already did that, I don’t remember lol
I had an English Bulldog (not neutered/raised from puppy) who was the kindest boy, and later, one who **was neutered** (adopted at age two) who unfortunately had major behavioral problems. I reckon the key differences in these two cases were primarily genetic and also the upbringing environment aspects. The latter dog was fearful easily and this typically indicates lack of exposure to various noises, people and situations in early puppy stages.
Great read, thanks for taking the time!
Could be multivariate, cutting across both biology/chemistry and the nurture aspects. Like for instance, how we deal with stress, how we learn to approach, communicate and solve questions. Learning how to have good outcomes to difficult conversations is a skill.
For instance, on the specific issue of violence between the sexes I've developed a nuanced take, whereby there's various contextual related issues at play, such as cultural norms (the stats of wives burnt alive in "kitchen accidents" in India for dowry related reasons are obscene), people snapping after years of nagging ("breaking point") and how our economic sphere breaks down dreams so that when reality shows up, men snap and vent frustration on the nearest person.
Correlations of T with psychological effects (especially sex drive) are nonlinear. Deprive a man of T and on average, his libido goes WAY down. (This happens most often in treatment of prostate cancer by Lupron.) But among men in the normal T range, there isn’t much of an association. If any.
Would you say this applies to exogenously administered testosterone? Because that seems to increase libido even among regular men.
I have tried to follow this literature, and that was not my impression. But it's not my exact area of expertise. Have a link/study?
I'm unclear on the point you're making about Androgen Insensitivity Syndrome? Do they not commit crime at far lower rates than normal males? My understanding was that chromosomal males with CAIS were behaviourally basically the same as women.
https://pubmed.ncbi.nlm.nih.gov/12710824/
This one doesn't specify criminality as being among the traits measured, but most other sex-differentiated psychological traits seem to point to those with CAIS being basically the same as women. Curious on your thoughts
My point was that it's not the testosterone itself that's to blame. People with CAIS have male-typical levels of testosterone, it just doesn't do anything. This makes it similar to the race comparison due to the SHBG point.
It's meant as kind of a reductio, but it's an ambiguous observation on its own.
Did you look at the impacts on people's behavior or subjective self-experience when they start taking testosterone? I've seen many claims of people feeling much more aggressive and violent when taking testosterone.
Second, is it possible there's some minimum level of testosterone, and so it's really the incremental testosterone that explains things (which would help explain the relatively small male-to-female testosterone gap, but the much larger male-to-female violent crime gap)? I know this was a timed post, so you probably didn't have room to go into that.
When your main argument is that we can't find strong evidence of testosterone leading to aggression, but we also don't know why men and women are so different in terms of violent crime, there's a Chesterton's fence sort of principle that makes me skeptical of your conclusion. I also know there is significant literature correlating in-utero testosterone exposure with more stereotypically male behavior though I'm far from an expert there.
I think there are also theories of high testosterone exposure leading to developmental changes that result in a high but comparatively smaller gap in resting or free testosterone after development.
re: 1. Yes, that is covered in this article.
re: 2. We should get samples of female bodybuilders for a proper test!
Was this the paragraph on 'roid rage? There seems to be a trend and it's even statistically significant so I think it's fair to say that testosterone increases aggression even if it may not be the primary driver of aggression.
Also, I believe that many people receive testosterone that is not a steroid but I haven't looked at that study's methodology.
A small thought and maybe totally off base, inspired by life on the farm.
Male cattle are castrated young (less than 90 days) but at one year steer and bulls are much closer together wrt to size than the steers are to heifers. After puberty, I think the difference is not as impressive between steer and cows, but the cows have way more body fat. The bulls of course are huge. I bet there is tons of evidence in the veterinary literature about the differences between intact males and, ahem, the minus 2 crew.
There is androgenizing effect that takes place during pregnancy. It seems to stop some time before birth, but it is enough to prime the brain. The phenomenon is known from mixed-sex calf twins where the female twin tends to develop into a freemartin, a frequently infertile chromosomally female calf. It is thought to be a combination of testosteron and anti-Müllerian Hormone (AMH).
Totally on board/familiar with the androgenizing effect in utero. (Did not know about the freemartin situation!) But I guess my focus here is how impressive post fetal growth is despite castration, and how much closer it is to typical male growth pre puberty. All seems like the growth and little boy, or little steer/male, antics are put into motion by something other than direct mediation of testosterone or already set in motion by birth.
Sorry, but the question of why men are more violent than women is completely different and unrelated to the question of why some men end up in jail, and you really only looked at evidence for the latter, as if it applied to your first question too, which it doesn't.
You should not be comparing testosterone levels between men to figure out why they're more violent than women, you should be comparing testosterone levels between men and women. Which men have about 15 to 45x as much as women, which would have been the much more obvious comparison in the chart you made regarding offender rates, and would seem to fully explain it. It would seem it is the jump from female ranges (typically under 70 ng/dL) to male ranges (265 to 900) that makes the difference, not the small differences in amounts once you're in male range.
As to comparisons between males rather than comparisons between sexes, I don't think criminal conviction is at all a good proxy for violence or aggression levels. In fact it's a very bad proxy. It's a much better proxy for having low intelligence and low social and actual capital.
There is lots of male aggression and violence that is not only not criminalized but celebrated. There are plenty of men smart enough not to get caught, or with the resources to stay out of jail if they do. Old men have more acquired capital to lose and young men don't, plus young men have more opportunities and access to sex with young women, so there's your reason for why they're more violent: they have more to win and less to lose.
I'm married to someone who has committed what might be described as extreme violence and has killed several people. Not only is he not a criminal, he is showered with awards and badges of honor for it. Bc he was a soldier. Plenty of men in MMA or other combat sports, actual combat, and plenty of other socially sanctioned activities that are not showing up in your numbers based on criminals who are mostly just dumb and lacking social resources.
The question of why men are more violent is not a mystery. Ask any evolutionary or behavioral biologists that focuses on animals. It's because the fundamental "purpose" if you will of males is just to fight other males for access to mating. Both sexes commit violence to protect and defend their territory, themselves, and (in the case of females mostly) their young. But when they're not doing those things or looking for food, pretty much the ONLY thing males do is fight with each other.
The question of why some men end up convicted for crimes and some do not is much more easily addressed by looking at how smart they are, what resources they have available to not be convicted, how entitled they've been taught to feel, what they have to lose, what they have to gain, and what their lethal and/or injurious capacity is. A guy who is likely to lose most fights probably won't get into them. A guy with a reputation and lots of assets to protect probably won't risk them. A young guy with no assets, no women, a high physical capacity for inflicting injury, and low IQ and self control...he probably will.
>You should be comparing testosterone levels between men and women. Which men have about 15 to 45x as much as women
He did that, noting the 17x higher levels in men.
>The question of why men are more violent is not a mystery. Ask any evolutionary or behavioral biologists that focuses on animals. It's because the fundamental "purpose" if you will of males is just to fight other males for access to mating.
That explains the reason why the trait evolved, not the neurochemical process by which it's imparted.
> That explains the reason why the trait evolved, not the neurochemical process by which it's imparted.
Indeed, a lot of people seem to be missing this. Yes, we can have sex differences in behavior without testosterone, and sex difference-related selection on adolescent and later behaviors can impact kids' behaviors. There are major sex differences in behavior in the mammal-expected direction among many invertebrate species, and they lack testosterone. It doesn't need to be invoked!
Would you say the reason why men are much hornier is because of testosterone?
It's a major part.
Related paper: https://www.tesble.com/10.1007/s10508-012-9946-2
I would assume so, given the drastic impact it has on women's libido when they take tiny itty bitty nothing amounts that are about a tenth or less of what men supplementing with it take.
Who actually cares about that though??
Are you actually taking issue with people having interest in things you're not interested in?
I just think you posed two questions at the beginning of the piece...and then answered different ones. None of this really looks at men as compared to women yet that is literally the opening line of the piece, so I assumed that's what it was about.
Hey, as a psychiatrist I was surprised by the "cure depression" part. Could you please elaborate ? I don't think we have such data.
Thank you for the read, was fun as always.
I believe you haven't cited “Sex hormone levels and body hair growth in !Kung San and Kavango men from Namibia,” by Eike-Meinrad Winkler & Kerrin Christiansen, American Journal of Physical Anthropology, Vol. 92, Issue 2 (Oct. 1993), pp. 155-164. Or a study showing that Sub-Saharan males have the highest testosterone levels of any race, A. Richard et al’s “Racial variation in sex steroid hormone concentration in black and white men: a meta‐analysis.” (2014) Andrology, 2(3), 428-435. I believe you also didn't cite See “Testosterone, Sex Hormone-Binding Globulin, and Body Composition in Young Adult African American and Caucasian Men,” by Stephen J. Winters et al, W.B. Saunders Company (2001). Forgive me if I did miss any of these studies and you did discuss them. I would be interested in knowing whether you think they are flawed in any specific sense: I've cited them all in my work and I certainly want to provide my own readers with a balanced assessment of their findings and what they imply.
As to why or whether men are more violent than women I don’t think that’s really the case. For example in unidirectional intimate partner violence women initiate it at about twice the rate as men, while in intimate partner homicide women commit it at roughly half the rate. People are conflating physical consequences with the decision to initiate it and it’s in the physical consequences that men have the lead.
The greater propensity for violence in men is clearly innate even from the most superficial life observations in childhood (before boys are more capable than girls at all) and adulthood, regardless of whether changes in testosterone can significantly alter aggression all by themselves. Without a background in all the relevant fields, I would simply guess that this arises from masculinized features of the brain that arise during sex differentiation.
How do these studies account for the testosterone-suppressive effects of opioids in the criminal population? Large proportion are using fentanyl plus other strong opiates and the men in prison know this, ask for T testing because they want to bulk up before a federal sentence. Older data would have heroin as opiate of choice. Violent behaviour persists with concomitant stimulant use and drug-induced psychosis, confounding T levels.
Hormones have both activational effects (at the time they are released) and organizational effects (that affect development and thus can persist after levels change). Sex hormones not only have activational effects but also organizational effects in utero, in infancy, and in puberty. Could you elaborate more on why you feel you ruled out prenatal effects? It looks like in your graph there was evidence for this via the digit ratios, yes? And it doesn't seem like you considered organizational effects during puberty or mini-puberty (the testosterone spike in infancy).
There is a sex difference in rough-and-tumble play in children as well as juvenile animals. In this period, there are not sex differences in androgens. However, androgens are implicated in this difference due to prenatal organizational effects. Girls with congenital adrenal hyperplasia have more rough-and-tumble play. Animal experiments have modified levels of rough-and-tumble play through manipulating androgen levels in utero. Yet if you tried to correlate rough-and-tumble play with testosterone at the time, you wouldn't find anything.
I'm not a statistician and I might have missed something critical in your post. If so, I apologize. But are we allowed to control for drug use if it's on the causal pathway?
+T -> increased reward seeing including drug use and sex -> violence ?
In *that case*, we shouldn't. But we've gotten around that through the use of many different designs.