With transgender people, the "original" state is unnatural. Medical treatment fixes this mismatch by making the body match the brain. It is a voluntary treatment with an extremely high success rate and negligible regret rates.
As far as I can tell, there is zero evidence that any remaining issues are caused by the treatment itself. All evidence points that the remaining issues are primarily socially related.
What did diabetics do before treatment and therapies? People with asthma, myopia, or really any other health issue? What does it matter?
> Do we know that the treatments help in the sense that treated people commit suicide less than untreated people before the treatments were available?
Why do we need to know that? What's the point of it? We don't evaluate effectiveness of any treatment by comparing current data with likely non-existent, pre-treatment (read: more than a century old, in this case) data.
There is clearly a significant social aspect here but I'm not willing to toss out that the body may just really not enjoy not being naturally in the expected configuration. Going to the gender you want seems to be partially solved but, like the person with the lost limb, it's doubtful we have anything close to a full substitute. I really worry the medical community has been doing a dis-service to the development of treatment by underplaying the suicidal aspects that may be less socially related.