How taking hormones impacts sexual preferences

Survey results

I asked trans people how much their preference for various sexual things changed after hormone replacement therapy! I detail exactly how I did this at the end. I’d also like feedback - peer review this for me, point out anything I did wrong, check my work!

But basically, I asked trans people how much HRT increased or decreased various sexual preferences, and looked at the difference in scores between estrogen and testosterone. I asked people to report their changes on a -3 (significantly decrease) to 3 (significant increase) scale of arousal, with 0 being no change.

Sample size was real low though, compared to my usual surveys - 60 people on testosterone, and 240 on estrogen. I used this to check correlations, but for absolute averages I filtered down to people who had been using regularly, for over 3 months, and this sample was 222 on estrogen, and 50 testosterone. I’d really like to get a higher sample size for this - if you or someone you know is on HRT, I’d love if you could take the survey - before you look at the results!

Results

I also asked about overall sex drive, and this was the biggest difference - on that same -3 to 3 scale, estrogen saw an average drop of -1.03, and testosterone a gain of 1.52.

This makes me a bit confused about the fact that on average, people on estrogen reported a 0.44 increase in interest in all the things I listed. It seems weird to say “my sex drive has dropped,” but then when asked “since taking estrogen, has your interest in [specific sexual interest] increased or decreased,” and in general you end up answering “increased.”

Maybe there’s some novelty here? Like, even if in general you feel less aroused, the novelty of having a newly changed sexual makeup causes you to report greater interest in things.

Or maybe this is an example of a reactionary sex drive; maybe the sex drive becomes less easily stimulated, but when it does get stimulated, it responds more intensely to things?

As for the gendered gaps in the data, it’s possible some or all of this is placebo or culturally informed - if you’re taking estrogen and identify as a woman, you might feel a lot of pressure to have standard ‘female’ sexual preferences. My guess is that this isn’t the case, at least not entirely, and that hormones are actually impacting sexual preferences, but my study isn’t designed in a way to determine this.

There are some interesting bits here though - the increased interest in trans people for both estrogen and testosterone, for example. This feels plausibly socially induced; maybe once you take hormones, you feel more solidly part of the trans community, which translates to greater sexual interest in trans people. Not sure though - I’m not on HRT, I’d be curious to hear theories about this from people who are!

Also, I know I already said this, but to reiterate, the sample size was pretty low, which makes me a lot more nervous about being susceptible to selection effects in unpredictable ways. If I get a better sample I’ll update this post!

Correlations

I only looked at correlations for trans women on estrogen, because my sample for testosterone was too small for me to bother. I’m not calculating odds ratios on this because I haven’t figured out how to do it in python yet. I’m paying attention to r values above around 0.27, because in my data this results in a p value at around 0.00005, which I can multiply by 200 or so and be safe.

• Nothing correlated significantly with age, length of time people have been on estrogen, or the regularity they’ve taken estrogen. Length of time people have taken estrogen and how old they are was correlated at 0.23, for calibration.

• Increase in sex drive was associated with a mild increase of interest in a lot of things, but especially interest in women, at 0.35

• Increase of interest in dominance was negatively correlated with increase of interest in submission, at -0.45, but correlated with increased interest in women at 0.28

• Increase of interest in submission was correlated with increase of interest in penises (r=0.49), nonconsent (as a receiver) (0.44), eagerness (0.41), bondage (0.37), trans people (0.31), power dynamics (0.29), and men (0.28).

• Increased interest in nonbinary people correlated with increased interest in trans people at r=0.65.

• Across all questions about changing sexual interest, the average correlation was r=0.19, which I interpret to mean that on average everything is correlated with everything a little bit, and if you wanna think about specific elements in contrast to other elements, you should adjust downward by 0.19

A lot of the sexual interests are correlated with other clusters of sexual interest. For example, interest in nonstandard ages is correlated with interest in incest, bestiality, brutality, etc.

I’ve been trying to study natural fetish clusters (and have a lot of data on this I’ll probably be publishing soon), but this is an unexpected potential pointer. It seems like HRT seems to pronounce specific clusters of fetishes for some people. I’m still a huge factor analysis noob, but I was like ‘what happens if I do a factor analysis on the correlations between the increase/decrease of interest?

I used hypothesize.io on default settings, and got what looked to be a ‘taboo’ type factor (brutality, bestiality, incest, nonstandard ages, creepiness, dirtiness, vore), a ‘feminine’ type factor (men, penises, being submissive, receiving noncon, eagerness, sensory stuff, gentleness), and a ‘standard kink’ factor (toys, bondage, bodily appearances/aesthetics, clothing, objects, gentleness, roles).

Then I was like, I have no idea how smart it is to do factor analysis on correlations, what happens if I do it on the data itself?

This attempt was way weaker, in that the factors seemed to explain less of the data, but the factors seem to be something like ‘rough receiver’ (receiving noncon, brutality, sadomasochism, humiliation, creepiness, power dynamics), ‘abnormal physical forms’ (mythical creatures, bestiality, nonstandard body parts, abnormal appearances), and ‘gender stuff’ (interest in trans, nonbinary, penises, my own gender identity, toys).

Anyway, all this is getting a bit off track, and I’m hoping to do a much more robust look into fetish clusters later on.

Methodology

You can find my exact questions and instructions test takers saw here, but basically I asked basic demographic stuff and if they were on estrogen or testosterone, how long they’d been on it, how regularly they’d been taking it, and then how much their sexual preferences for [long list of things] had changed since taking HRT, on a 7-point spectrum of “this is more arousing” to “this is less arousing”.

I didn’t ask about absolute interest, only change.

I posted a link to the survey on twitter, but at some point a bunch of radical trans activists, somehow convinced that I was doing this for nefarious purposes, started encouraging people to take my survey and fill in bad data. I took down my survey and closely scoured all responses during the time period this happened, aggressively removing any response sets that looked slightly suspicious. I checked the remainder against the results that had come in before the brigade, and found it was still roughly consistent, and so kept them. The remainder currently makes up about 23% of my current dataset.

I also posted the link to the survey on fetlife, in various discord chats, and on my facebook, and asked people to share the survey with any trans people they know.

I only looked at people who were sort of classically trans (e.g., assigned male at birth, identifies not as a man, is taking estrogen), and removed responses that featured e.g. a female taking estrogen.

You can find the raw data, averages, correlations, and confidence intervals here. I did a little bit more whittling down in that dataset before pulling the numbers from it. The red rows are ones that came from the brigaded time period.