A Bayesian network meta-analysis of 178 randomised trials (81,014 adults) estimated placebo-adjusted weight loss at 52 weeks. Retatrutide 12 mg led at -19.80 percentage points (95% CrI -21.81 to -17.78), followed by tirzepatide 15 mg (-16.87) and cagrilintide-semaglutide 2.4/2.4 mg (-16.37). Only tirzepatide 15 mg and semaglutide 2.4 mg had high-certainty evidence. Thirteen of 32 strategies raised adverse-event discontinuation versus placebo. People with type 2 diabetes lost 62% as much weight as those without.
The headline numbers matter less than the evidence gaps the authors expose. Of 160 target treatment strategies, only 45 had real 52-week estimates; the rest rely on extrapolation. For a patient and clinician choosing a starting regimen, that means most dose and titration decisions rest on indirect inference rather than head-to-head trials. The retatrutide figure comes from a drug not yet widely approved, with limited trial data, so its ranking carries wider uncertainty than the certainty grades for the two established agents. The diabetes attenuation finding is practically useful: expectations should be set lower for this group. Weight loss is a surrogate, and this analysis does not address mortality or cardiovascular events, where semaglutide and tirzepatide have separate outcome trials. This is a preprint and has not been peer reviewed, so estimates and certainty ratings may shift. Its main contribution is methodological and diagnostic, a confirmatory map of what is known and unknown, rather than a change in the efficacy hierarchy already suggested by individual trials.