Orforglipron Worked Across Menopause Stages, With One Asterisk

Menopause is one of the few moments in adult life when weight gain is close to expected. Falling estrogen shifts fat toward the abdomen and, for many women, makes it harder to lose. So when Eli Lilly reported that its daily GLP-1 pill orforglipron (sold as Foundayo) produced similar weight loss in women before, during and after menopause, the claim landed on a real question. The data were first shown at the American Diabetes Association meeting in June 2026 and presented again at the EASD meeting in Milan, which ran from September 28 to October 2.

The headline holds up in one of the two trials. In the other, it rests on groups too small to carry much weight.

Where the numbers come from

The analysis reused two 72 week, placebo controlled phase 3 trials. ATTAIN-1 enrolled adults with obesity, or overweight plus a weight related condition, who did not have diabetes. ATTAIN-2 enrolled people with obesity or overweight and type 2 diabetes. Lilly's researchers pulled out the female participants, more than 1,500 in total, and grouped them as premenopausal, perimenopausal or postmenopausal. All authors of the EASD abstract are Lilly employees.

In ATTAIN-1, women on the highest dose lost 12.8% of their body weight if premenopausal (171 women), 14.4% if perimenopausal (142) and 14.1% if postmenopausal (152). In every group, more than 80% lost at least 5% of their weight, and between 40.9% and 51.5% lost at least 15%. Waist circumference fell by 11.4 to 12.5 centimeters. The EASD summary puts the placebo comparison at up to 83% of treated women reaching 5% weight loss, against 30% on placebo.

Those figures sit within less than two points of one another. If menopause blunted the drug's effect, this is where it would show, and it does not.

The diabetes trial is a different story

In ATTAIN-2, the averages spread out: 11.3% weight loss before menopause, 8.9% during perimenopause and 13.6% after. The share losing at least 15% ranged from 20.4% in perimenopausal women to 44.2% after menopause.

Then look at the group sizes in Lilly's own table for the highest dose. The premenopausal group in ATTAIN-2 contains 10 women. The perimenopausal group contains 33. Only the postmenopausal group, at 109, is reasonably sized. A gap of nearly five points between perimenopause and postmenopause could be real, could reflect who happened to land in each group, or could be noise. With 10 and 33 women, there is no way to tell.

That matters because perimenopause is precisely the window many women with diabetes ask about. The honest reading of ATTAIN-2 is not that the drug works less well in perimenopause. It is that this analysis cannot answer the question for women with diabetes.

Why post hoc matters here

A post hoc analysis slices a trial after the fact along lines it was not designed to test. Randomization balances the treatment and placebo arms overall, but not necessarily each slice. Menopausal stage also overlaps heavily with age, so these groups differ in more than hormones. None of that makes the ATTAIN-1 result wrong. It means the result describes consistency, which is useful, rather than proving that menopause has no effect on the drug, which the analysis cannot do.

The context is commercial as well as clinical. Orforglipron competes in an oral GLP-1 market where differentiation sells, and our coverage of its head to head trial against oral semaglutide and its comparison with insulin shows Lilly building that case trial by trial. A clean message about menopause is valuable to a company. That does not make it false, but it explains why the presentation emphasized the similarity and not the small groups.

What would settle it

A prospective trial in women stratified by menopausal stage from the start, with enough perimenopausal women with diabetes to compare, and with body composition measured, would answer the question properly. So would a pooled analysis across several GLP-1 drugs, which would at least multiply the sample sizes.

For women without diabetes, this is genuinely reassuring data: the drug did not appear to lose its effect around menopause, the stage when weight is hardest to move. For women with diabetes in perimenopause, the 8.9% figure should be read as a placeholder, not a prediction. Ten women cannot define an average for millions, and the most useful thing this analysis does is show exactly where the next trial needs to recruit.