Design: Phase 3, double-blind, randomised, placebo-controlled, four arms
Population: 238 Japanese adults with obesity disease (BMI 27 or above with obesity-related disorders), about 25% with type 2 diabetes
Primary outcome: Percentage change in bodyweight and proportion losing at least 5% at week 72
Effect: −5.9% (6 mg), −8.7% (12 mg), −7.9% (36 mg) vs −0.6% placebo; ≥5% loss in 46–61% vs 15%

ATTAIN-J randomised 238 Japanese adults with obesity disease to once-daily oral orforglipron 6, 12 or 36 mg, or placebo, for 72 weeks. Entry followed Japanese criteria: a BMI of 27 to 35 with two obesity-related disorders, or 35 and above with one, and about a quarter had type 2 diabetes. It was published in Lancet Diabetes & Endocrinology on 30 October and funded by the manufacturer.

Weight fell by 5.9% on 6 mg, 8.7% on 12 mg and 7.9% on 36 mg, against 0.6% on placebo. About six in ten on the two higher doses lost at least 5% of their weight, against 15% on placebo. Stopping because of nausea was most common on 36 mg (8%).

The weight loss is smaller than the injectable incretins achieve, but this is an oral, small-molecule drug rather than an injected peptide. In this population the highest dose bought more side effects and no more weight loss, which is a useful dosing signal.

The trial does not report its regulatory status in India; check the current CDSCO position before discussing it with patients.

Expect about 6 to 9% weight loss at 72 weeks from oral orforglipron in East Asian adults, less than with injectable incretins.
In this trial 12 mg did as well as 36 mg, with fewer stops for nausea.
Asian BMI thresholds for obesity disease are lower; the trial enrolled from a BMI of 27.
Gastrointestinal effects were mostly mild to moderate, as with other GLP-1 receptor agonists.

Why it matters

An oral, food-independent incretin could widen access where injections and cold chains are barriers, but its effect size sits below the injectables.

The statistics, in plain English

The 12 mg group lost 8.7% (95% CI 6.5 to 10.9) against 0.6% on placebo. The confidence intervals for 12 mg and 36 mg overlap almost completely, so the trial cannot show that the higher dose works better. Each group had about 60 people, so the ranges are wide.