A pill that works as well as an injection is the thing most readers want to exist, and the trial evidence now goes some way toward it. OASIS 4 randomized 307 adults with overweight or obesity and no diabetes, two to one, to oral semaglutide 25 mg once daily or placebo for 71 weeks [1]. The result is real and the caveats are specific, and both matter before you compare what any seller lists as an oral.
The result
Estimated mean change in body weight from baseline to week 64 was −13.6 per cent with oral semaglutide and −2.2 per cent with placebo, an estimated difference of 11.4 percentage points (95% CI −13.9 to −9.0, P < 0.001). Participants on the drug were significantly more likely to reach reductions of 5, 10, 15 and 20 per cent or more, and to report improved physical function on the IWQOL-Lite-CT scale.
Set against the pooled injectable figure of about −10.7 per cent from the Cochrane review, this is not a lesser product. It is a different trial in a different population, so the two numbers should not be subtracted from one another, but the pill is clearly in the same range.
The gastrointestinal cost
Gastrointestinal adverse events occurred in 74.0 per cent of the oral semaglutide group against 42.2 per cent on placebo. That is a high rate in absolute terms and a wide gap against placebo, and it is the number most likely to describe your first three months. What the trial reports is frequency, not severity; the practical shape of it is covered in titration and nausea.
What you are buying is probably not this
This is the part a seller’s page will not make clear. OASIS 4 tested a specific manufactured 25 mg tablet. Most oral semaglutide sold through telehealth is compounded — a sublingual drop or a dissolving tablet prepared by a pharmacy, often at a dose the seller does not publish. The FDA does not review a compounded preparation before it is dispensed, and no compounded oral product has been through a trial like this one.
Two practical consequences. An oral price cannot be set against an injection price as though the milligrams meant the same thing. And a seller that publishes no dose beside an oral figure has published a number you cannot interpret at all — which is common enough that we counted it across the market in what sellers will not tell you.
What the trial does not settle
Three hundred and seven participants is a small trial by the standards of this drug class, it ran 71 weeks, and it was funded by the manufacturer. It says nothing about what happens after stopping, and nothing about compounded preparations at all.