There's no nationwide research infrastructure for gleaning insights like that. But large provider organizations have huge patient populations and do a lot of research. They have multiple studies underway for Ozempic and similar drugs.

That's a really good point, and it seems to pervade the scientific establishment. There's an unfortunate tendency to study something, and (assuming a real effect is found & reproduced) say 'well we know about that now' and just move onto something else, instead of building out infrastructure to gather more longitudinal/geographic data. The efforts in the UK to build huge datasets for long-term diagnostics is a welcome exception.