Retatrutide took off about a fifth of body weight in trials, more than any other drug in a new head-to-head ranking, and it is not even approved yet.
That number is a 22.10 percent average weight loss versus placebo (95 percent confidence interval 25.60 to 18.60 percent). It comes from a network meta-analysis published August 27 in BMJ Medicine ↗ by Chen and colleagues at the First Hospital of China Medical University in Shenyang. A network meta-analysis pools many separate trials and uses their shared comparisons, mostly against placebo, to line up drugs that were rarely tested head-to-head. This one swept up 58 randomised trials and 24,214 adults who were overweight or had obesity but did not have diabetes.
The ranking put the newer, multi-target drugs on top. Retatrutide ↗, Lilly's once-weekly injection that hits three gut-hormone receptors at once, led at roughly 22 percent. Tirzepatide ↗, sold as Mounjaro and Zepbound, came next at 19.28 percent, a bit over 20 pounds off a 250-pound starting weight. CagriSema, Novo Nordisk's combination of the amylin drug cagrilintide ↗ and semaglutide ↗, landed third at 17.32 percent. The older single-receptor drugs, semaglutide and liraglutide used on their own, came in more modest. Waist circumference and blood lipids followed the same order.
The staircase is softer than it looks
The tidy ranking is less solid than a list of percentages suggests. The authors call it a probabilistic hierarchy, which means the order is a bet about which drug is most likely best, not a measured gap between them, and the confidence intervals for several drugs overlap. Very few of these trials pitted one active drug against another. Almost all of the comparisons run through placebo, which is exactly what a network meta-analysis is built to stitch together, but also where it is weakest. The authors graded the overall certainty of the evidence as low. So "retatrutide beats tirzepatide" is the direction the data lean, not a settled result.
Tolerability told a different story than efficacy. Low-certainty evidence pointed to higher treatment-discontinuation rates for retatrutide and for danuglipron, Pfizer's oral non-peptide GLP-1 pill. Meanwhile mazdutide ↗, a GLP-1 and glucagon dual agonist, looked better tolerated than its weight-loss figure alone would predict. The most effective drug and one of the hardest to stay on can be the same drug. That trade-off is the thing a single ranking line hides.
What the order is really sorting by
The split the ranking draws is mostly by receptor count. The conventional drugs work through the GLP-1 receptor ↗ alone. The ones on top add more levers. Tirzepatide and retatrutide also pull the GIP receptor ↗, retatrutide and mazdutide add the glucagon receptor ↗, and CagriSema pairs GLP-1 with amylin. More hormone systems engaged, more weight off, and for at least two of them, more people walking away.
The order here echoes what a real-world cohort found in July, when All of Us records ranked five marketed GLP-1 shots and put tirzepatide first ↗. What this analysis adds is the top of the list. Retatrutide, CagriSema, mazdutide and danuglipron are mostly still in trials, so the ranking reaches past the pharmacy shelf to drugs a patient cannot fill yet. Retatrutide's headline number is consistent with its own Phase 3, which cleared 28 percent weight loss at the top dose ↗ with an 11 percent dropout. The ceiling of this class is still moving up. Whether patients can live at that ceiling is the question the tolerability column keeps asking.