The Rhythm | Issue 20
Issue 20 | Week of 17th August 2026
No major cardiology congress fell this week. The strongest evidence came from The Lancet Healthy Longevity, JACC's special issue on obesity, JAMA Cardiology, and the European Heart Journal.
Cardiology Intelligence Briefing: the trouble with average
Cardiovascular evidence keeps hitting the same wall this week: broad categories work, until they don't fit the patient in front of you. Take age. A trial of statin discontinuation in the very old asked whether stopping was worse than continuing. It wasn't, in a carefully selected group of adults 75 and older, though the trial was too small to turn that answer into a rule.1 Take weight. In a quarter of a million people, body mass index compressed something as basic as body composition into a single number, and missed real cardiovascular risk hiding in how fat is actually distributed.3 Take mechanism. An adjunctive therapy for ST-segment elevation myocardial infarction (STEMI) looked biologically persuasive on paper. Tested against hard outcomes in an unselected population, it did not significantly improve its primary endpoint.2 Three settings, one lesson: the average patient is rarer than the category assumes.
Precision isn't only about splitting populations more finely. Sometimes it's about who generated the evidence in the first place. A small serial-imaging substudy found encouraging signs of cardiac remodelling, but only in a carefully selected subset of a larger amyloidosis trial; the clearest disease-modification signals often come from the patients easiest to follow closely, not the ones most representative of routine care.4 A Swedish national study flipped that problem. Patients with an established diagnosis of Alzheimer's disease are routinely left out of anticoagulation trials altogether, so an important clinical question has only ever been answered by observational data, with all the uncertainty that brings.5
None of this week's papers hands clinicians a new default. What they do is sharpen the edges of the defaults already in use. The lesson isn't that cardiovascular practice needs more categories. It's that the categories already in use shouldn't be mistaken for the patients inside them.
