Every intake form in this market asks for a height and a weight, and somewhere behind it sits a number that decides whether you are offered anything at all. That number is not a fact about bodies. It is a guideline, and guidelines differ — which is a different kind of gate from the state lists sellers publish but works the same way.
A trial run at lower numbers
STEP 12 enrolled adults across 19 sites in mainland China and Taiwan using locally defined BMI thresholds: 24 to under 28 with at least one weight-related condition, or 28 to under 30 with or without type 2 diabetes. [1] Of 242 participants, 161 received semaglutide 2.4 mg weekly and 81 placebo, for 44 weeks, alongside lifestyle intervention. Half were women and 19.4% had type 2 diabetes.
Those entry criteria sit below where the well-known trials began. The rationale is not novel and it is not this desk’s: East Asian guidelines set overweight and obesity cut-points lower because cardiometabolic risk rises at lower BMI in these populations. The trial was designed around the local definition rather than the American one.
What it found
Body weight fell 12.1% on semaglutide against 2.2% on placebo — an estimated treatment difference of 9.9 percentage points, 95% CI −11.8 to −8.0, p<0.0001. The proportion reaching at least 5% weight loss was 80.5% against 24.4%, an odds ratio of 14.8, 95% CI 7.4 to 29.6.
Adverse events were reported by 141 of 161 participants on the drug, 87.6%, and 61 of 81 on placebo, 75.3%. Both figures are high, and the difference between them is smaller than the drug figure alone suggests — which is the reason to always ask what the placebo arm reported.
How it compares
The 12.1% reduction is smaller than the headline figures from the large Western trials, and the honest comparison is not straightforward. This cohort started at a lower BMI, so there was less weight available to lose, and the trial ran 44 weeks rather than 68. Weight loss expressed as a percentage behaves differently depending on where you start, which is one reason expected-loss estimates should never be read as a promise.
What does travel is the shape: a large, clearly significant separation from placebo on both co-primary endpoints, consistent with what the wider trial evidence shows. A drug that works only in the populations it was first tested in would be the surprising result, and this is not that.
Why a single trial in one region matters here
Because most readers meet this drug through a corpus of trials conducted largely in the United States and Europe, and the generalization outward is usually assumed rather than demonstrated. Trials like this one are how the assumption gets tested, and they tend to arrive years later and attract a fraction of the attention.
It is also a reminder that the published evidence is narrower than the marketing implies at any given moment — the same gap that separates a famous trial from the literature behind it. The trial was funded by Novo Nordisk, which makes the drug.