GLP-1s Go Cardio: What a New Meta-Analysis Says About Heart Attack Risk

Number needed to treat is the unglamorous metric that separates a press release from a real drug. It answers: how many people have to take this thing, for how long, to prevent one bad event? The meta-analysis ran that calculation across an average follow-up of about three and a half years and landed on figures that are, by cardiology standards, genuinely good.

To prevent one heart attack, 207 patients need to be treated. To prevent one major adverse cardiovascular event — the composite of MI, stroke and cardiovascular death that regulators care about — the NNT drops to 67. Cardiovascular mortality alone: 170. Stroke: 335. For context, statins in primary prevention typically post NNTs in the same neighborhood for MACE over similar timeframes. GLP-1RAs are now playing in that league, and they are doing it in a population that often cannot tolerate or fully benefit from statins alone.

The meta-analysis flagged a meaningful effect modifier: higher BMI was associated with greater MI risk reduction in GLP-1RA users (β: -0.09, p = 0.03). Translated: the further someone sits from a healthy body composition, the more cardiac upside the drug appears to deliver. This is consistent with the biology — these molecules act on weight, glycemia, blood pressure and inflammation simultaneously — but it is also a useful filter for the busy 40-year-old asking whether this conversation applies to him.

If you are lean, metabolically healthy and have no family history of premature ASCVD, the absolute benefit on display in these trials is small. If you carry significant visceral adiposity, prediabetes or type 2 diabetes, or established cardiovascular risk, the same relative risk reduction translates into a much larger absolute payoff. The drug is not a generic longevity tool. It is a cardiometabolic intervention whose value tracks your starting risk.

Importantly, the cardiovascular benefit held across patients with and without type 2 diabetes. That matters. It moves GLP-1RAs out of the diabetes silo and into a broader cardioprotective conversation — one that primary-care physicians and cardiologists are now having in real time.

Standard lipid panels are a blunt instrument. Two men with identical LDL-C numbers can carry very different atherosclerotic risk depending on the size and density of their lipoprotein particles — small, dense LDL is the troublemaker. A 52-week randomized trial published this year went looking at that finer-grained picture in patients with type 2 diabetes.

The trial randomized 34 obese T2D patients to weekly subcutaneous semaglutide or daily oral sitagliptin, with 31 non-diabetic obese controls. After a year, semaglutide produced significant reductions in BMI, waist circumference and HbA1c, alongside lower LDL cholesterol and non-HDL cholesterol and a redistribution of LDL and HDL subfractions toward a less atherogenic profile. Sitagliptin moved glycemia modestly and barely touched the lipid composition.

The mechanistic kicker: multivariate regression suggested the lipoprotein shifts were not simply downstream of weight loss or glycemic improvement. Something else — a direct pleiotropic effect of GLP-1 signaling on lipid handling — appears to be doing work. The sample is small and the population specific, so caution is warranted before extrapolating. But it is a coherent piece of evidence alongside the meta-analysis: better lipid quality, not just a lower number on the scale.

None of this is free. The same meta-analysis quantified the cost: GLP-1RAs roughly increased gastrointestinal side effects by 55% versus placebo, with a number-needed-to-harm of 9 for GI events. Nausea, reflux, constipation, the early-titration tax everyone on these drugs knows about. For most patients these symptoms attenuate, but they are not rare and they are the main reason people quit.

What the trial data does not tell you is how the math shifts at lower, off-label or recreational doses, or in lean people chasing modest cosmetic weight loss. The MACE numbers come from patients with elevated cardiovascular risk treated to therapeutic doses for years. Anyone considering a GLP-1RA — particularly for cardiovascular prevention rather than diabetes — should have that conversation with a clinician who can weigh personal risk against the trade-offs honestly.

For the reader sizing this up against squat numbers and resting heart rate: the meta-analysis does not make GLP-1RAs a default biohack. It makes them a serious cardiovascular drug for people with serious cardiovascular risk. If your waist circumference, fasting glucose, ApoB or family history puts you in that bucket, the case is now strong enough to put on a clinician's desk and discuss explicitly — alongside lifestyle work that no injectable replaces. If you are metabolically tidy and chasing the last five pounds, the risk-benefit math from these trials simply does not apply to you.

The bigger shift is conceptual. We are watching a class of peptides graduate from weight-loss drugs to cardiometabolic drugs, with the trial evidence to back the new label. That is the rarer kind of medical news: not a hype cycle, but a recalibration.