The risk that doesn't stay in the pelvis
Twenty-five years of records from nearly three million women, and one number worth carrying into your next appointment: a 46% higher risk of type 2 diabetes among women with endometriosis. The association was strongest in the women least likely to be screened for it.
Before anything else
This is an association found in medical records, not a diagnosis and not a prediction about you. GLP-1 medications are not a treatment for endometriosis and nothing below suggests otherwise — we are writing about metabolic risk, which is a different subject that happens to sit in the same body. Bring it to a clinician who knows your history.
On August 6 a study landed in Diabetologia that we have been turning over ever since. Researchers on the ARCHES project — Advancing Research on Cardiovascular Health and Endometriosis — went through twenty-five years of medical records in the Utah Population Database, 1996 through 2021, covering nearly three million women. Roughly 100,000 of them had a diagnosis of endometriosis.
Those women were 46% more likely to develop type 2 diabetes.
The work was led by Maggie Fuzak Nunziato, a doctoral student in epidemiology at George Mason, with Anna Pollack and collaborators at the University of Utah, the University of Arizona, Brigham and Women's Hospital, and Intermountain Health. It is a retrospective cohort study, which means it can show a pattern and cannot show a cause. We want to be clear about that before we say anything else about it.
The part that changes the conversation
A 46% figure on its own is the kind of statistic that gets a headline and then gets forgotten. The subgroups are what made us sit up.
The association was strongest among premenopausal women. It was strongest among women without obesity. And it was strongest among women with extra-pelvic endometriosis, where lesions are found outside the pelvic cavity entirely.
Read that again through the eyes of a screening protocol. Premenopausal, not heavy, no obvious metabolic flag — that is precisely the woman a standard diabetes risk calculator waves through. She does not get the fasting glucose. She does not get the A1c. If she mentions fatigue, it gets filed under the endometriosis, because everything gets filed under the endometriosis.
The women at highest risk in this data are the women least likely to be tested for it.
Why this belongs on a comparison site
We compare GLP-1 providers. We are not going to pretend that a diabetes-risk study is a reason to start a medication, and we would be suspicious of anyone who framed it that way.
What it is, is more evidence for the thing this entire site is built around: conditions that get labelled reproductive are frequently metabolic too, and care that splits them into separate departments will miss things. We made the same argument about PMOS (PCOS) in You are not a willpower problem and about the early GLP-1 receptor findings in The clue in the peritoneal fluid. This is the same shape of argument with a much larger dataset behind it.
It also raises the stakes on something we grade. If you have endometriosis and you are working with a telehealth provider on weight or metabolic health, whether that provider ever looks at a metabolic panel — rather than a scale and a dose schedule — is no longer a nice-to-have. The Care Standard weights clinical depth and monitoring for exactly this reason.
What to do with it
Not much, and that is honest. One study does not change screening guidelines, and it should not send anyone into a spiral. What it does give you is a reason to ask — and a citation to point at if the asking gets brushed off.
Some questions that tend to land well. Given my endometriosis history, is there a case for checking my fasting glucose and A1c even though I don't fit the usual risk profile? If we're going to monitor, what's the interval — annually, or only if something else changes? Is there anything in my history, like extra-pelvic disease, that would make you want to look sooner? And who is holding this — you, or my gynaecologist, or nobody?
That last one is the question underneath all of it. Fragmented care is how a 46% signal goes unnoticed in an individual woman for a decade. Someone has to be looking at the whole of you.
What we're watching
ARCHES is a cardiovascular-health project first, so we expect more from this group on heart risk in endometriosis, and we will cover it the same way — carefully, with the study design named out loud. There is also a broader reframing underway in the field; the Endometriosis 2026: A Nerve-Centric Disease meeting in New York this year put the nervous system at the center of the disease rather than the pelvis. Different mechanism, same direction of travel: endometriosis is turning out to be a whole-body condition that we have been treating as a local one.
None of which shortens anyone's diagnostic wait. But it does mean the next generation of women asking these questions may be asking them of clinicians who already expect the answer to be complicated.
The honest disclaimer. This is general information from the women who write Selene — not medical advice, not a recommendation, and not a prediction about your results. The study described here is observational and shows association, not causation. GLP-1 medications are not a treatment for endometriosis; references to endometriosis on this site describe whole-woman clinical context, never a therapy for the condition. No GLP-1 is FDA-approved specifically for PMOS (PCOS). Selene is editorial: we do not prescribe, sell, or dispense anything.
Sources. Fuzak Nunziato M, Pollack A, et al., Diabetologia, published 6 August 2026 (ARCHES cohort, Utah Population Database); George Mason University College of Public Health announcement, August 2026; Society for Women's Health Research, Endometriosis 2026: A Nerve-Centric Disease, 2026.