Large Utah Study Links Endometriosis With Higher Type 2 Diabetes Risk

An August 6 study in Diabetologia linked a documented endometriosis diagnosis with a higher rate of later type 2 diabetes diagnosis. The adjusted hazard ratio was 1.46, with a 95 percent confidence interval from 1.43 to 1.50.

That is a population-level association. It does not show that endometriosis causes diabetes, it does not predict one person's risk, and it does not create a new diabetes-screening recommendation.

How the study was built

The researchers assembled a dynamic cohort of 2,939,364 people assigned female at birth in the Utah Population Database from 1996 through 2021. They identified endometriosis and type 2 diabetes from ICD-9 and ICD-10 diagnosis codes. During a mean follow-up of 10.8 years, 99,978 people in the cohort were diagnosed with endometriosis.

The analysis treated endometriosis as a time-varying exposure and adjusted for birth year, birth state, race and ethnicity, age, and BMI at cohort entry. The association varied by endometriosis subtype and metabolic context. George Mason University's August 10 release highlighted that it appeared stronger in some groups not traditionally viewed as having the highest baseline diabetes risk.

A cohort this large can detect patterns that smaller studies may miss. Size alone does not turn an observational records study into a causal experiment.

What the 46 percent figure means

A 46 percent higher adjusted rate is relative. It does not mean that 46 additional people out of every 100 will develop type 2 diabetes, and it is not an absolute-risk estimate for an individual reader. The study compared diagnosis rates over time after statistical adjustment.

The result is notable partly because prior studies had generally reported little or no overall association. One large new analysis can change the evidence conversation, but it does not settle it.

The limits matter

The authors specifically caution about residual confounding and detection bias. People managing endometriosis may have more contact with the health system, creating more opportunities for both conditions to be recorded. Diagnosis codes also cannot capture every difference in symptom severity, care access, treatment, or metabolic history.

Those limits leave several explanations open. Shared inflammatory or metabolic pathways may contribute, unmeasured factors may influence both conditions, or part of the signal may come from differences in who gets diagnosed and when. The study was not designed to choose among those explanations.

Where lab testing fits

Endometriosis is not a diabetes test, and this paper does not say that everyone with endometriosis needs testing on a new schedule. Established blood markers still have distinct jobs. An A1c test reflects average blood sugar over roughly two to three months, while a fasting glucose test measures blood sugar at one fasting time point.

LTS previously covered how fasting-glucose trajectories tracked with later cardiometabolic risk. Both studies are reminders that population patterns can identify questions worth studying without diagnosing an individual or replacing clinical guidance.

The bottom line

The ARCHES study is a meaningful signal because of its scale, long observation window, and explicit subtype analysis. Its responsible interpretation is still narrow: diagnosed endometriosis was associated with higher type 2 diabetes diagnosis rates in this Utah cohort. Replication, more detailed clinical data, and guideline review would be needed before endometriosis alone changes diabetes-screening practice.

This article is editorial commentary and is not medical advice. It has not been reviewed by a physician and should not be used to make decisions about diabetes screening, diagnosis, or testing frequency.