On serious events, they stop looking alike, and the answer depends on which outcome you pick. The pooled analysis below is where the two drugs separate.
Direction can also flip with the comparator. In veterans with type 2 diabetes and opioid use disorder, overdose risk fell against two comparator drugs and not against three [2]. So the answer changed with what the drug was set beside. That is true of most claims in this class.
This site already covers the largest single cohort comparing these two drugs. It found tirzepatide ahead on weight and most of each group gone before the end. This is the pooled version, and it adds one finding the cohort did not report.
The efficacy side, briefly
Ten studies, randomized and observational, with at least 24 weeks of follow-up, covering 41,381 participants. [1] Tirzepatide produced 4.28 percentage points more weight reduction, 95% CI −5.28 to −3.28. Absolute loss ran 4.43 kg greater, 95% CI −5.56 to −3.30. Glycated hemoglobin fell 0.29 percentage points further.
It was likelier to get people past 10%, 15% and 20% weight loss. At the 5% threshold there was no difference. Both drugs get most people that far. The separation is entirely in the deeper end.
Pooling randomized trials with observational studies invites an obvious objection. The authors addressed it. Subgroup analyses by study design produced consistent findings, as did subgroups by diabetes status.
The finding that is new here
Three safety comparisons came back level. Overall adverse events were similar. Gastrointestinal adverse events were similar. Discontinuation because of adverse events was similar, at a risk ratio of 1.28 with p=0.54.
Serious adverse events were not. They were more frequent with tirzepatide, risk ratio 1.83, p=0.007.
Why the pattern is the interesting part
If tirzepatide were simply harder to tolerate, it should show up in the ordinary adverse events and in people quitting. It did not. The two drugs looked alike on everything except the most severe category.
That could mean several things. It could reflect the greater metabolic effect of the stronger drug producing more events at the tail. It could be a consequence of who receives which drug in the observational studies. It could be a chance finding in a category with few events. That is exactly why the missing interval matters, and why a signal without its uncertainty is so hard to use.
What a reader should take from it
The efficacy gap is well established and this adds to it. Tirzepatide takes off more weight, and the advantage is concentrated among people aiming past ten per cent rather than at modest loss.
The safety gap is a question rather than an answer, and it is one nobody selling either drug will raise. A seller comparing the two has every reason to quote the weight difference and none to mention a serious adverse event ratio. and almost nothing about safety appears there at all. Where the two drugs genuinely look alike, that is worth knowing too, because similar is not the same as identical. Unlike most comparisons in this field, these studies put the two drugs against each other directly.