Literally none of that supports your hypothesis, and you haven't established the first factor: are testosterone treatments statistically significant within the target demographic?