Continuous glucose monitor data on semaglutide — what worked for you?

roxy_nash said:

HbA1c reflects roughly three months of average glycaemia weighted toward the most recent weeks, which is why repeating it at six weeks tells you very…

All true, with one condition: that curve is for people who reached the dose on schedule. Anyone who slowed the ladder for tolerability is on a different, flatter curve, and comparing yourself with the published mean will make you feel like a non-responder when you are not.

TrialNerd_Beth said:

The mechanism that matters here is not stomach emptying, it is central.

That reframing is the part I needed. Sending this to two other people who asked me the same thing last week.

roxy_nash said:

HbA1c reflects roughly three months of average glycaemia weighted toward the most recent weeks, which is why repeating it at six weeks tells you very…

I read this differently from roxy_nash, on substance rather than tone. The trial means are being read too generously in this thread. STEP populations were selected, supported, and titrated by protocol, and the real-world curves are consistently a few points worse. That difference is not noise, it is what happens when you remove the study infrastructure.

One concrete data point for the thread. Because it is glucose-dependent, this class carries a low intrinsic hypoglycaemia risk on its own — the risk arrives when it is combined with insulin or a sulfonylurea, which usually need reducing.

Moderator note: the sourcing question belongs in the vendor section and has been split out.