It's better to have too much detail than too little. At least it's trivial to strip away data if you want, but we can't add it back in if it wasn't there to begin with.

Also, it seems that sexual orientation has different spreads across genders (judging by other studies), which is why I added in gender.