The Dose That Stopped Helping

ACHIEVE-5 tested three doses of oral orforglipron on top of insulin in 546 adults with long-standing type 2 diabetes. Blood sugar improved at every dose — and the top dose was not better than the middle one.^[1] Tripling from 12 mg to 36 mg bought nothing on HbA1c, which is not what the word “higher” leads people to expect.

The middle row

Every dose beat placebo at P<0.001, and the confidence intervals for 12 mg and 36 mg overlap almost entirely. On blood sugar, the effect has reached a ceiling somewhere between 3 and 12 mg, and going higher does not move it.

The two endpoints stop responding at different doses.

Weight kept falling as the dose rose — 2.6%, then 4.8%, then 5.4%. So the same drug shows a plateau on one outcome and a continuing gradient on another. Anyone reasoning that a higher dose is simply a stronger version of the same thing has the wrong model, and the dose that is right depends on which problem is being treated.

Who was in this trial

That context changes how the numbers should be read. We recently covered a trial of a different drug in people managing diabetes with diet and exercise alone, a median of 2.5 years from diagnosis, which posted an HbA1c reduction of 1.94% — retatrutide’s phase 3. The figures look comparable. The populations are not remotely comparable, and we are not setting them against each other.

Achieving nearly two points of HbA1c reduction on top of titrated insulin, fourteen years in, is a harder thing than achieving the same reduction in someone recently diagnosed and taking nothing.

Why the weight number looks small

A 5.4% reduction at the top dose is far below what orforglipron produced in obesity trials, and that is expected rather than disappointing. Insulin causes weight gain — the placebo group here gained 0.2% while having their insulin titrated upward. The drug was working against that.

Reading this figure against an obesity trial would be a category error of the kind this register documents regularly. Different population, different concurrent treatment, different question. Orforglipron’s obesity results are covered in ATTAIN-2.

What it is good for

• It answers a real clinical question. Someone on insulin whose blood sugar is still too high has limited options, and adding an oral drug rather than more injections is a meaningful one.
• Completion was high — 507 of 546, or 92.9% — which is a good sign for tolerability in a 40-week trial.
• The ceiling is practically useful. If 12 mg does what 36 mg does for blood sugar, the higher dose is worth taking only for the extra weight effect, which is a different trade to weigh.
• Forty weeks is short for a condition managed over decades, and this trial reports no cardiovascular or kidney outcomes.