Research methods
Weight-loss percentage vs cardiometabolic outcomes
Why a larger percent body-weight reduction in a trial does not automatically translate into greater cardiovascular, kidney, hepatic or mortality benefit.
The biggest weight-reduction headline number is not automatically the best outcome.
Headline weight numbers in obesity trials
Tirzepatide produced larger mean weight reduction than semaglutide 2.4 mg in SURMOUNT-5; phase 2 retatrutide produced still larger numbers. These are real, but they are obesity-trial weight numbers, not outcome numbers.
Outcome trials remain the deciding evidence
SELECT showed semaglutide reduced MACE in adults with overweight or obesity and pre-existing CVD without diabetes. SURPASS-CVOT showed non-inferiority of tirzepatide vs dulaglutide for MACE-3 in T2D+ASCVD. SUMMIT showed tirzepatide benefit in HFpEF+obesity. FLOW showed semaglutide benefit in T2D+CKD. Outcome benefit is established trial-by-trial in defined populations, not transferred automatically between drugs because one drug has a higher weight number.
Editorial caution
Claiming "X is therefore the best" based on a weight-reduction comparison alone is not evidence-led. The fair frame is: X had a larger weight-reduction effect in the studied trial; outcome benefit is established where outcome trials have been conducted; cross-class comparison requires careful methodology.
Key takeaways
- Larger weight reduction in a trial is not equivalent to greater cardiovascular, kidney or mortality benefit.
- Outcome benefit is established drug-by-drug, indication-by-indication.
- Editorial framing should track the evidence, not the headline number.
- Formal statistical training.
- Individual treatment, prescribing or dosing advice.
- Permission to skip the original publication of any trial cited.
Sources are tagged by class. Regulatory documents, peer-reviewed publications and trial registry records are preferred. Manufacturer disclosures and conference material are accepted where peer-reviewed publication is not yet available; preprints are flagged.