The weight-loss pill arrived: what actually changed

For as long as these medicines have been in the public conversation, the honest answer to "is there a pill?" was no — not for weight. There was an oral GLP-1, Rybelsus, but its licence covered type 2 diabetes, and every treatment licensed for weight management was something you injected.

That changed across nine months, in three steps.

This is the part that gets lost when both are reported as "the weight-loss pill". They are different molecules, and what that buys is not a different percentage but a different daily life.

Semaglutide is a peptide, which the digestive tract is built to destroy. Getting it absorbed from a tablet at all requires an absorption enhancer and a set of conditions, and those conditions are written into the licence: swallow it on an empty stomach after at least eight hours without food, with no more than half a glass of water — up to 120 ml — and eat, drink and take nothing else by mouth for at least thirty minutes afterwards. Eating sooner reduces how much gets in. One tablet a day, and never two to make a bigger dose.

Orforglipron is a small molecule, so none of that applies. The licence sets no time of day, no fasting requirement and no limit on water. You swallow it whole, with or without food, whenever you remember.

For someone who skips breakfast anyway, the semaglutide routine is close to invisible. For someone who wakes up and has tea, or who takes a thyroid tablet or a blood-pressure tablet first thing — and the thirty-minute rule covers other oral medicines too — it is a daily negotiation that will eventually be lost. The licences do not rank the two medicines, and neither should anyone else; the question is which set of rules fits a real morning.

Both figures below come from each medicine’s own Summary of Product Characteristics rather than from a press release, and both are the result in everyone randomised rather than only in those who completed.

The temptation is to line them up and conclude that the semaglutide tablet does what the semaglutide injection does — 13.6% against 13.7% — and that both fall well short of tirzepatide. The first half of that is a cross-trial comparison, and this blog has a whole post arguing that cross-trial comparisons are unreliable. It would be poor form to abandon the argument the moment it produced a convenient answer.

The populations were not the same. OASIS 4 enrolled 307 people over 64 weeks; the semaglutide arm of SURMOUNT-5 was part of 751 people over 72 weeks. Trial length, starting weight, the intensity of the diet and activity programme and the handling of people who stopped early all differ, and each of those moves a mean by more than the gap between 13.6 and 13.7.

What can be said without stretching is narrower and more useful. Each tablet beat its own placebo by about 11 percentage points. Roughly a quarter of people on the semaglutide tablet and a fifth on orforglipron lost at least a fifth of their body weight, which is a scale of result that did not exist in tablet form two years ago. And the licence itself treats the tablet as a real alternative rather than a lesser one: it allows someone on 2.4 mg of the injection to move across to 25 mg tablets, starting a week after their last injection.

A tablet removes three obstacles that have nothing to do with how well a medicine works.

The needle, first, which is a smaller barrier than people assume once they have tried it and a complete one for those who never will. Then the cold chain: pens live in a fridge, have an in-use clock running against them and complicate every holiday, and a tablet does none of that. And manufacturing — peptide synthesis capacity has been a real constraint on supply in this category, while a small molecule is made by ordinary chemistry. Shortages have kept more people off treatment than side effects have.

What it does not change is the hard part. A tablet does not make the gastrointestinal effects milder, because those come from the mechanism and not from the route: nausea, constipation, diarrhoea and indigestion are the common side effects of both tablets, exactly as they are of the injections. Orforglipron adds a consideration the peptides do not have — it is metabolised like an ordinary medicine, so its dose is capped at 9 mg alongside a strong CYP3A4 inhibitor or an OATP1B inhibitor, which makes the list of your other medicines a real part of the conversation.

And appetite still returns when treatment stops, on a tablet exactly as on an injection. The thing that holds a result is what gets built while the medicine is making it easier, which is the one part of this that no approval will ever shorten.

One consequence of the headlines deserves stating plainly: "there is a weight-loss pill now" is being used to sell things that are not it.

There is no licensed liquid GLP-1 for weight management anywhere, and there is no licensed patch, gummy or drop. Every authorised GLP-1 receptor agonist is an injection or one of the tablets above, and all of them are prescription-only. So the capsules and "liquid GLP-1" on shop shelves and marketplace listings are not weaker versions of what was just approved; they contain no GLP-1 receptor agonist at all and have been through no assessment of effectiveness or safety.

The same applies to powders sold as "research grade" orforglipron. Those listings existed before the medicine was licensed and they still exist, with the difference that they now imitate something a pharmacy can legitimately dispense — which makes them more convincing and no safer. A licensed medicine reaches you one way: an assessment, a named prescriber, and a registered pharmacy.