Guidelines recommend these drugs after a heart attack or stroke in people with type 2 diabetes. A national registry can check whether that happens. It mostly does not.
The number that does not need a causal claim
Czech registry data from 2015 to 2024 identified 126,845 people surviving a first nonfatal heart attack, of whom 28,206 had confirmed type 2 diabetes, and 177,115 surviving an ischemic stroke, of whom 73,750 had it. [1] Roughly 2% of those survivors were started on a GLP-1 within twelve months. Use was lowest among women and older adults.
That finding stands whatever you conclude about effectiveness. A recommended therapy reached one eligible patient in fifty. Nothing about study design changes it, and it is a different kind of gap from the access limits sellers publish.
What the matched comparison found
Users were propensity-matched to non-users. After heart attack, the matched cohort was 2,271 people, 401 of them on a GLP-1, followed a median of 35 months. Major adverse cardiovascular events came in at HR 0.70, 95% CI 0.52 to 0.93. All-cause death was HR 0.61, 95% CI 0.47 to 0.80. Cardiovascular death was HR 0.54, 95% CI 0.36 to 0.80.
After stroke, the matched cohort was 2,235 people, 385 on a GLP-1, median follow-up 27 months. MACE was HR 0.71, 95% CI 0.54 to 0.94. All-cause death HR 0.59, 95% CI 0.46 to 0.76. Cardiovascular death HR 0.55, 95% CI 0.37 to 0.81.
How much to believe
The direction is almost certainly right. Randomized trials already established that these drugs reduce cardiovascular events in people with diabetes and existing disease, and those trials are the reason to believe it. This registry is consistent with them.
The magnitude is another matter. Four hundred treated patients, matched out of tens of thousands, with a mortality benefit that spills past the mechanism, is not a number to quote as an effect size. It is the same caution that applies to any matched-records analysis, and it applies harder here because the treated group is small.
Why a two percent figure matters here
This site sells nothing to people recovering from a heart attack, and none of the clinical findings above transfer to a weight-loss purchase. The transferable part is about supply and attention. In a country with universal coverage and a recommending guideline, 98% of eligible patients did not get the drug — while a large commercial market sells the same molecules to people at far lower risk.
That is not hypocrisy on anyone’s part and it is worth noticing. The gap between who a drug is proven to help and who actually buys it is the widest fact in this market, and it is invisible from inside a trial. What sellers record about who they ship to is thin, measurably so.