Every comparison you have read between these two drugs until recently was, technically speaking, invalid.
Not dishonest — just structurally incapable of supporting the conclusion people drew from it. Putting the 15 percent from one trial next to the 21 percent from another and declaring a winner ignores that the two studies enrolled different people, ran for different lengths, and applied different background diet and activity programmes.
What was actually compared
Adults with obesity and without type 2 diabetes, randomised to tirzepatide or semaglutide, each escalated to their maximum tolerated dose, followed for 72 weeks. Same protocol, same population, same clock.
That design is what makes the result quotable. It removes the excuse that the difference was an artefact of who enrolled.
~20%
Mean weight reduction, tirzepatide, at 72 weeks
SURMOUNT-5
~14%
Mean weight reduction, semaglutide, same trial
SURMOUNT-5
72 wks
Duration — both arms at maximum tolerated dose
Why this changes less than the headline suggests
Three reasons the result should not reorganise your treatment.
Averages conceal individuals. A six-point gap between group means says almost nothing about any single patient. Plenty of people lose more than 20 percent on semaglutide. Plenty lose under 10 percent on tirzepatide. The distributions overlap enormously, and there is no test that tells you in advance which side of them you are on.
Coverage usually decides first. For a large share of Americans this is settled by a formulary before a clinician forms a view. The most effective drug you cannot obtain is worse than the second most effective drug you can.
Weight is not the only endpoint. Semaglutide carries the longer track record and the broader set of completed outcome trials — cardiovascular, kidney, liver. For a patient whose central worry is a heart attack rather than a dress size, depth of evidence is a legitimate reason to choose the drug that lost the weight comparison.
A trial tells you what happened to a group. It does not tell you what will happen to you, and the gap between those two sentences is where most bad decisions in this field live.
Where it genuinely is decisive
Two situations.
If you are choosing between the two with no coverage constraint, no relevant comorbidity, and weight reduction as the primary goal — the evidence now points one way, and it is reasonable to act on it.
And if you have completed a fair trial of semaglutide at a maximum tolerated dose, given it a year, and landed well short of what you and your clinician were aiming at, this result strengthens the case for trying the other molecule rather than concluding that drugs do not work for you.
That switch is much easier to make well with a record of the first attempt behind you — doses, dates, weights and side effects, rather than a recollection of a year you were not documenting. Any tracker does this; Zenday keeps the three on one timeline, which is the format the conversation actually needs.
The thing to watch
Head-to-head trials are expensive and commercially risky, which is why they are rare. This one happened because a company was confident enough of the answer to pay for it.
That confidence is worth noticing when the next generation arrives. If the makers of the newer agents run direct comparisons, they expect to win. If they publish only against placebo, draw your own conclusion about what they think a direct comparison would show.