5 ms·

If you want to do the work I can recommend a book called The Trouble With Testosterone that takes a meta perspective on the different research on the topic. It

by belorn 6y ago

If you want to do the work I can recommend a book called The Trouble With Testosterone that takes a meta perspective on the different research on the topic. It is a fun evening/day going through it cited studies. > The data we have from chemical castration of sex offenders is very solid The data show same result for female chemical castration of sex offenders. Usually the study goes something like this: a self selected set of child sex offenders volunteer for chemical castration, and later they report a diminished desire for sex. The problem with linking this to aggression is on several layers. It self selected for individuals that want to reduce their desire for sex. It is self reported. And it is about diminished desire for sex. Will reducing bodily strength cause a reduction in aggressiveness? If we leave the topic of sex offenders and look at aggressiveness as a whole, bodily strength might very well effect social status and thus fit the challenge hypothesis. Crippling a high social status individual and they will likely spend less energy on defending their status, through the result is going to vary depending on the culture. The empathy studies are interesting (the book above list a few of those). According to one such study, giving men increased testosterone does not decrease their ability to read emotion in an eye test, but giving it to women does reduce their ability. Again this show a repeating pattern of those studies where within normal range, high or low testosterone levels can not predict behavior. You need to give people testosterone so that they are outside human levels to get any significant effect, and then the result is often minor unless taken to a very extreme level. > As far as theoretical arguments against it go, I'd say the strongest is that high levels are simply a proxy for some irreversible virilization that happened earlier in life. Which would much better explain why both criminal male and females are relatively high T. What is your source for that? Criminals have higer rate of violence, and if violence causes testosterone secretion, then the higher level can be explain by the increase violence. male-male competition in prison is also likely a bit more intense than outside which would also explain a higher level of testosterone.

auganov 6y ago
> The data show same result for female chemical castration of sex offenders As expected. Why would it not work in females? You'll be easily getting these levels to a very "unnatural" level below the low end for both males and females. The studies aren't just on self reported feelings. You're looking at reductions in recidivism exceeding 50%. Yes some studies show a much smaller reduction, but AFAIK every study does show some improvement. Sure as I noted myself it could be self-selection as some argue, but nonetheless there isn't a study out there outright disproving it. There are a few countries where it's mandatory but not sure if there's any data from these. When you've got studies showing efficacy, to disprove them you'd want something of higher rigor. Say a proper RCT. If one says self selection taints the result then you want a study that controls for that. Again I'm absolutely not opposed to the notion that there's some cultural/societal feedback loop involved. I'm just arguing that administering strong blockers MAY help with many issues relating to aggressiveness. And that these arguments on just how linear or predictable the relationship is are interesting but not that relevant to disproving the practical efficacy of such a treatment. I fully agree that one shouldn't demonize the hormone itself and anyone who has high levels. That's what that line of research tells me as far as I'm concerned. > What is your source for that? Regardless of what causes violence we have a huge undeniable discrepancy between males and females. If we go with the explanation that violent behavior bumps T up that goes nowhere in explaining the enormous sex difference. Don't remember the exact ratio off top of my head by it's definitely somewhere around 1 to 10 so not buying a purely societal explanation. And if you look at the extremes, say serial killers or very brutal forms of murder it grows even wider apart. So it would seem logical to me that in these violent high T females either the T does matter or it doesn't but is an indication of virilization earlier in life. As to the claim that testosterone levels would be a proxy for earlier virilization - for example digit ratios are believed to be an indicator of virilization in utero and studies do show that they correlate with adulthood T levels[0]. Pretty sure there are studies relating it to criminality too. Personally my hunch is that it's a bit of both. Say if we put everyone on these drugs I'd imagine overall violent crime would drop but males would still be vastly outnumber females. [0] https://europepmc.org/article/med/21592297 https://europepmc.org/article/med/21592297
- belorn 6y ago
  Just because castration can have an effect on recidivism for some forms of sex offenders (and please point towards a source because it does make the discussion more worth having), it does not mean it have an effect on say axe murderers. Gun control may also have an effect on recidivism for gang related violence involving military weapons, but is unlikely to effect child sex rapists. Not everyone that work for once set of criminals work for an other set of criminals, nor are recidivism and aggressiveness the same thing. And there are studies where violent criminals of the non-sex offending nature were given testosterone reducing drugs. It did not work. > Regardless of what causes violence we have a huge undeniable discrepancy between males and females That kind of research are plenty in gender research. Men are X more likely to be math professors, having a job, and a number of other gender stereotypical aspect. There is a lot of culture involved and it very hard to separate culture from biology. But even considering what we know about biology, aggression and crime are not the same and female on female aggression tend to be different than male on male aggression. Culturally we tend to ignore the former and focus on the later. And finally, there is some evidence that in many cultures male aggression get rewarded with social status, and in the challenge hypothesis, that would link aggression with testosterone through culture. Game theory experiment has demonstrated that testosterone can influence players to be more or less aggressive in an economic game depending on the design of the rules. If aggression is rewarded with maintaining status then injecting players with testosterone increase aggression. If the game is changed and it is cooperation that is rewarded, players with raised testosterone play more cooperative. All cited in that above book.
  - auganov 6y ago
    Quoting from Heim & Hursch (1979): "Of 102 sex offenders, only three castrates (2.9%) relapsed into sex offenses. However, in only 41 cases was the period of observation long (5-10 years). Therefore, there was a maximum of 7% recidivism during this observation period. The recidivism incidence before castration was as follows: Of 102 subjects, 59 (58%) had relapsed." 2.9%/7% vs 58% (this one is on surgical on chemical) Fedoroff, J.P., Wisner-Carlson, R., Dean, S., & Berlin, F.S. (1992): "All patients received equivalent amounts of group psychotherapy. Of these, 17/46 (37%) relapsed. The rate of relapse among subjects receiving treatment with medroxy-progesterone acetate (MPA) was 4 out of 27 subjects (15%) whereas the rate of relapse among subjects not receiving MPA was 13/19 (68%) "[1] 15% vs 68% And since you bring up math professors, there's actually a study showing a single dose of testosterone improved spatial cognition in females [2]. Again I'm not opposed to cultural angles. I don't doubt that culture makes small innate differences look much bigger. But that's more philosophical/anthropological, it doesn't tell us much on whether a drug would or wouldn't work at scale. We could go on and deconstruct every psychiatric drug this way. [0] https://doi.org/10.1007/BF01541244 https://doi.org/10.1007/BF01541244 [1] https://doi.org/10.1300/J076v18n03_10 https://doi.org/10.1300/J076v18n03_10 [2] https://www.sciencedirect.com/science/article/pii/S0166432815302680 https://www.sciencedirect.com/science/article/pii/S016643281...

by belorn 6y ago

auganov 6y ago

> The data show same result for female chemical castration of sex offenders As expected. Why would it not work in females? You'll be easily getting these levels to a very "unnatural" level below the low end for both males and females. The studies aren't just on self reported feelings. You're looking at reductions in recidivism exceeding 50%. Yes some studies show a much smaller reduction, but AFAIK every study does show some improvement. Sure as I noted myself it could be self-selection as some argue, but nonetheless there isn't a study out there outright disproving it. There are a few countries where it's mandatory but not sure if there's any data from these. When you've got studies showing efficacy, to disprove them you'd want something of higher rigor. Say a proper RCT. If one says self selection taints the result then you want a study that controls for that. Again I'm absolutely not opposed to the notion that there's some cultural/societal feedback loop involved. I'm just arguing that administering strong blockers MAY help with many issues relating to aggressiveness. And that these arguments on just how linear or predictable the relationship is are interesting but not that relevant to disproving the practical efficacy of such a treatment. I fully agree that one shouldn't demonize the hormone itself and anyone who has high levels. That's what that line of research tells me as far as I'm concerned. > What is your source for that? Regardless of what causes violence we have a huge undeniable discrepancy between males and females. If we go with the explanation that violent behavior bumps T up that goes nowhere in explaining the enormous sex difference. Don't remember the exact ratio off top of my head by it's definitely somewhere around 1 to 10 so not buying a purely societal explanation. And if you look at the extremes, say serial killers or very brutal forms of murder it grows even wider apart. So it would seem logical to me that in these violent high T females either the T does matter or it doesn't but is an indication of virilization earlier in life. As to the claim that testosterone levels would be a proxy for earlier virilization - for example digit ratios are believed to be an indicator of virilization in utero and studies do show that they correlate with adulthood T levels[0]. Pretty sure there are studies relating it to criminality too. Personally my hunch is that it's a bit of both. Say if we put everyone on these drugs I'd imagine overall violent crime would drop but males would still be vastly outnumber females. [0] https://europepmc.org/article/med/21592297 https://europepmc.org/article/med/21592297

belorn 6y ago
Just because castration can have an effect on recidivism for some forms of sex offenders (and please point towards a source because it does make the discussion more worth having), it does not mean it have an effect on say axe murderers. Gun control may also have an effect on recidivism for gang related violence involving military weapons, but is unlikely to effect child sex rapists. Not everyone that work for once set of criminals work for an other set of criminals, nor are recidivism and aggressiveness the same thing. And there are studies where violent criminals of the non-sex offending nature were given testosterone reducing drugs. It did not work. > Regardless of what causes violence we have a huge undeniable discrepancy between males and females That kind of research are plenty in gender research. Men are X more likely to be math professors, having a job, and a number of other gender stereotypical aspect. There is a lot of culture involved and it very hard to separate culture from biology. But even considering what we know about biology, aggression and crime are not the same and female on female aggression tend to be different than male on male aggression. Culturally we tend to ignore the former and focus on the later. And finally, there is some evidence that in many cultures male aggression get rewarded with social status, and in the challenge hypothesis, that would link aggression with testosterone through culture. Game theory experiment has demonstrated that testosterone can influence players to be more or less aggressive in an economic game depending on the design of the rules. If aggression is rewarded with maintaining status then injecting players with testosterone increase aggression. If the game is changed and it is cooperation that is rewarded, players with raised testosterone play more cooperative. All cited in that above book.
- auganov 6y ago
  Quoting from Heim & Hursch (1979): "Of 102 sex offenders, only three castrates (2.9%) relapsed into sex offenses. However, in only 41 cases was the period of observation long (5-10 years). Therefore, there was a maximum of 7% recidivism during this observation period. The recidivism incidence before castration was as follows: Of 102 subjects, 59 (58%) had relapsed." 2.9%/7% vs 58% (this one is on surgical on chemical) Fedoroff, J.P., Wisner-Carlson, R., Dean, S., & Berlin, F.S. (1992): "All patients received equivalent amounts of group psychotherapy. Of these, 17/46 (37%) relapsed. The rate of relapse among subjects receiving treatment with medroxy-progesterone acetate (MPA) was 4 out of 27 subjects (15%) whereas the rate of relapse among subjects not receiving MPA was 13/19 (68%) "[1] 15% vs 68% And since you bring up math professors, there's actually a study showing a single dose of testosterone improved spatial cognition in females [2]. Again I'm not opposed to cultural angles. I don't doubt that culture makes small innate differences look much bigger. But that's more philosophical/anthropological, it doesn't tell us much on whether a drug would or wouldn't work at scale. We could go on and deconstruct every psychiatric drug this way. [0] https://doi.org/10.1007/BF01541244 https://doi.org/10.1007/BF01541244 [1] https://doi.org/10.1300/J076v18n03_10 https://doi.org/10.1300/J076v18n03_10 [2] https://www.sciencedirect.com/science/article/pii/S0166432815302680 https://www.sciencedirect.com/science/article/pii/S016643281...