SGLT2 Inhibitors and Mortality in Endometriosis With Type 2 Diabetes: A Propensity-Matched Analysis

We conducted a retrospective, propensity score–matched cohort study using deidentified electronic health record data from the TriNetX federated research network, which includes more than 130 million patients from 102 health care organizations. Adults with endometriosis and T2D were identified and classified according to treatment with SGLT2i (predominantly in combination with metformin) or metformin-based therapy without SGLT2i. Propensity score matching (1:1) balanced demographic, clinical, laboratory, and medication variables (standardized mean differences <10%). The primary outcome was all-cause mortality; secondary outcomes included MACE and progression to chronic kidney disease (CKD) stages 4 to 5. All-cause mortality in TriNetX was ascertained using structured electronic health record fields indicating deceased status and, when applicable, supplemented by International Classification of Diseases, Tenth Revision (ICD-10) mortality codes, in accordance with TriNetX outcome definitions. Time-to-event analyses were performed using a cause-specific Cox proportional hazards model, with time zero defined as the cohort-specific index date and death treated as the event of interest; competing risks methods were not applicable for an all-cause mortality endpoint. Follow-up continued until death or censoring, with censoring defined as the earliest of: 1) the date of the last recorded health care encounter (loss to follow-up); or 2) the end of the prespecified analytic time window (administrative censoring). Censoring did not reflect treatment discontinuation or switching; medication changes during follow-up were not captured as censoring events, and treatment persistence and switching could not be modeled as time-varying exposures. HRs with 95% CIs were estimated using Cox proportional hazards models. Analyses were performed within the TriNetX platform. Detailed methods, the participant flow diagram, and variable definitions are provided in the Supplemental Appendix .

Among patients with endometriosis and T2D, 7 . 334 received an SGLT2i and 11 . 338 received metformin as part of metformin-based therapy without SGLT2i before propensity score matching ( Table 1 , Central Illustration ). Baseline metformin use was common in the SGLT2i cohort. The cohort overall had a high burden of cardiovascular risk factors and obesity (body mass index, mean [SD], 35.8 [8.5] and 36.3 [8.7] kg/m 2 ; overweight/obesity, 55.3% and 47.6%; hypertension, 73.2% and 59.8%; dyslipidemia, 72.7% and 52.5%). Compared with metformin users, SGLT2i users were older (54.2 [11.2] vs 50.9 [11.3] years) and had greater cardiometabolic comorbidity, including higher prevalences of dyslipidemia (72.7% vs 52.5%; standardized difference, 42.8%), hypertension (73.2% vs 59.8%; 28.7%), kidney disease (20.7% vs 7.3%; 39.5%), coronary artery disease (17.9% vs 7.0%; 33.4%), and HF (16.1% vs 3.6%; 42.8%). Use of cardiovascular medications was also more frequent among SGLT2i users. Table 1 Baseline Characteristics of Adults With Endometriosis and Type 2 Diabetes Treated With SGLT2 Inhibitors or Metformin, Before and After Propensity Score Matching Before Propensity Score Matching After Propensity Score Matching SGLT2i (n = 7,334) Metformin (n = 11,338) Std Diff. (%) SGLT2i (n = 3,072) Metformin (n = 3,072) Std Diff. (%) Demographics  Age (y), mean ± SD 54.2 ± 11.2 50.9 ± 11.3 29 52.8 ± 10.8 53.0 ± 11.1 1.7  White, n (%) 3,694 (50.4%) 6,174 (54.5%) 8.2 1,526 (49.7%) 1,517 (49.4%) 0.6  Black or African American, n (%) 1851 (25.2%) 2,462 (21.7%) 8.3 667 (21.7%) 674 (21.9%) 0.6  Asian, n (%) 660 (9%) 1,007 (8.9%) 0.4 325 (10.6%) 341 (11.1%) 1.7  Hispanic or Latino, n (%) 799 (10.9%) 1,388 (12.2%) 4.2 381 (12.4%) 353 (11.5%) 2.8  Socioeconomic and psychosocial circumstances, n (%) 462 (6.3%) 408 (3.6%) 12.5 157 (5.1%) 145 (4.7%) 1.8 Risk factors  Hypertension, n (%) 5,369 (73.2%) 6,779 (59.8%) 28.7 2083 (67.8%) 2084 (67.8%) 0.1  Smoker, n (%) 1,050 (14.3%) 1,253 (11.1%) 9.8 362 (11.8%) 339 (11%) 2.4  Overweight or obesity, n (%) 4,056 (55.3%) 5,402 (47.6%) 15.4 1,564 (50.9%) 1,511 (49.2%) 3.5  Dyslipidemia, n (%) 5,332 (72.7%) 5,947 (52.5%) 42.8 2058 (67%) 2043 (66.5%) 1  Alcohol-related diagnoses, n (%) 87 (1.2%) 145 (1.3%) 0.8 37 (1.2%) 44 (1.4%) 2 Cardiovascular comorbidities  Heart failure, n (%) 1,182 (16.1%) 412 (3.6%) 42.8 244 (7.9%) 236 (7.7%) 1  Coronary artery disease, n (%) 1,312 (17.9%) 794 (7%) 33.4 340 (11.1%) 337 (11%) 0.3  Myocardial infarction, n (%) 355 (4.8%) 158 (1.4%) 19.9 75 (2.4%) 70 (2.3%) 1.1  Ischemic stroke, n (%) 334 (4.6%) 300 (2.6%) 10.3 115 (3.7%) 116 (3.8%) 0.2  Intracranial hemorrhage, n (%) 30 (0.4%) 18 (0.2%) 4.7 10 (0.3%) 10 (0.3%) 0  Atrial fibrillation or flutter, n (%) 486 (6.6%) 295 (2.6%) 19.3 116 (3.8%) 109 (3.5%) 1.2 Noncardiovascular comorbidities  Kidney disease, n (%) 1,520 (20.7%) 826 (7.3%) 39.5 382 (12.4%) 392 (12.8%) 1  COPD, n (%) 558 (7.6%) 482 (4.3%) 14.3 169 (5.5%) 167 (5.4%) 0.3  Sleep apnea syndrome, n (%) 1944 (26.5%) 1961 (17.3%) 22.4 658 (21.4%) 676 (22%) 1.4  Peripheral vascular disease, n (%) 322 (4.4%) 188 (1.7%) 16 70 (2.3%) 61 (2%) 2  Previous cancer, n (%) 3,554 (48.5%) 5,281 (46.6%) 3.8 1,473 (47.9%) 1,506 (49%) 2.1  Cognitive impairment, n (%) 10 (0.1%) 10 (0.1%) 1.4 10 (0.3%) 10 (0.3%) 0 Laboratory tests and examinations  Systolic BP (mm Hg), mean ± SD 128.7 ± 17.4 128.0 ± 17.2 4.2 128.5 ± 16.6 128.4 ± 17.9 0.8  Diastolic BP (mm Hg), mean ± SD 76.9 ± 11.5 76.9 ± 11.8 0.1 77.6 ± 10.9 76.7 ± 11.9 7.9  Heart rate, mean ± SD 83.2 ± 14.1 82.6 ± 14.3 4.7 83.0 ± 13.8 83.0 ± 14.5 0.2  Body mass index (kg/m 2 ), mean ± SD 35.8 ± 8.5 36.3 ± 8.7 6.6 35.5 ± 8.3 35.7 ± 8.7 2.4  Total cholesterol (mg/dL), mean ± SD 174.9 ± 46.7 189.9 ± 46.1 32.3 181.0 ± 48.2 183.1 ± 51.4 4.1  LDL cholesterol (mg/dL), mean ± SD 95.2 ± 38.3 109.9 ± 37.5 39 100.1 ± 39.0 102.5 ± 39.3 6  HDL cholesterol (mg/dL), mean ± SD 45.6 ± 15.4 46.1 ± 15.7 3.2 46.3 ± 14.9 45.1 ± 16.0 7.7  Triglyceride (mg/dL), mean ± SD 180.6 ± 139.6 175.9 ± 148.1 3.3 186.1 ± 148.3 185.9 ± 185.6 0.1  Estimated GFR (MDRD, mL/min), mean ± SD 82.0 ± 29.4 88.4 ± 26.7 22.7 87.1 ± 28.0 86.7 ± 27.9 1.5  HbA1c (%), mean ± SD 8.1 ± 1.9 7.5 ± 1.9 36 8.0 ± 1.8 8.1 ± 2.1 5 Medications  Beta-blockers, n (%) 3,313 (45.2%) 3,637 (32.1%) 27.1 1,212 (39.5%) 1,203 (39.2%) 0.6  Calcium-channel blockers, n (%) 2,475 (33.7%) 2,379 (21%) 28.9 861 (28%) 893 (29.1%) 2.3  ACE inhibitors, n (%) 2,292 (31.3%) 2,642 (23.3%) 17.9 914 (29.8%) 933 (30.4%) 1.3  Angiotensin II inhibitors, n (%) 2,500 (34.1%) 2017 (17.8%) 37.8 868 (28.3%) 883 (28.7%) 1.1  ARNI, n (%) 199 (2.7%) 10 (0.1%) 22.5 10 (0.3%) 10 (0.3%) 0  MRA, n (%) 989 (13.5%) 690 (6.1%) 25.1 260 (8.5%) 261 (8.5%) 0.1  Diuretics, n (%) 3,362 (45.8%) 3,705 (32.7%) 27.2 1,194 (38.9%) 1,206 (39.3%) 0.8  Digitalis glycosides, n (%) 92 (1.3%) 42 (0.4%) 9.9 21 (0.7%) 22 (0.7%) 0.4  Amiodarone, n (%) 131 (1.8%) 45 (0.4%) 13.4 27 (0.9%) 23 (0.7%) 1.4  Lipid-lowering drugs, n (%) 5,218 (71.1%) 4,867 (42.9%) 59.5 1996 (65%) 2020 (65.8%) 1.6  Metformin, n (%) 4,963 (67.7%) 11,338 (100%) 97.7 3,063 (99.7%) 3,072 (100%) 7.7  Sulfonylureas, n (%) 1833 (25%) 1,027 (9.1%) 43.4 567 (18.5%) 561 (18.3%) 0.5  GLP-1 receptor agonists, n (%) 2,544 (34.7%) 960 (8.5%) 67.2 644 (21%) 671 (21.8%) 2.1  DPP4 inhibitors, n (%) 1,427 (19.5%) 640 (5.6%) 42.6 435 (14.2%) 440 (14.3%) 0.5  Thiazolidinediones, n (%) 391 (5.3%) 175 (1.5%) 20.9 105 (3.4%) 114 (3.7%) 1.6  Insulin, n (%) 1778 (24.2%) 858 (7.6%) 46.8 465 (15.1%) 447 (14.6%) 1.6  Antiplatelet therapy, n (%) 2,223 (30.3%) 2,101 (18.5%) 27.7 747 (24.3%) 742 (24.2%) 0.4  Anticoagulant, n (%) 477 (6.5%) 349 (3.1%) 16.1 134 (4.4%) 127 (4.1%) 1.1 ACE = angiotensin-converting enzyme; ARNI = angiotensin receptor–neprilysin inhibitor; BP = blood pressure; COPD = chronic obstructive pulmonary disease; DPP4 = dipeptidyl peptidase-4 inhibitors; GFR = glomerular filtration rate; GLP-1, glucagon-like peptide-1; HbA1c = hemoglobin A1c; HDL = high-density lipoprotein; LDL = low-density lipoprotein; MDRD = Modification of Diet in Renal Disease; MRA = mineralocorticoid receptor antagonist; SGLT2i = sodium glucose co-transporter 2 inhibitor; Std diff. = standardized difference. Central Illustration SGLT2is + Metformin vs Metformin Cardiovascular outcomes in endometriosis patients endometriosis affects ∼10% of reproductive-age women (∼190 million worldwide) and is characterized by ectopic proliferation of endometrial-like tissue outside the uterine cavity. Emerging data suggest increased atherosclerotic cardiovascular disease risk in affected women. In this retrospective, propensity score–matched cohort study using the TriNetX federated research network, adults with endometriosis and type 2 diabetes were identified and classified by treatment with sodium glucose co-transporter 2 inhibitors plus background metformin vs metformin-based therapy without sodium glucose co-transporter 2 inhibitors. After matching, 6.144 patients (3.072 per group) were followed for 2.8 ± 1.8 years, median 2.8, IQR: 3.7. All-cause mortality was significantly lower with sodium glucose co-transporter 2 inhibitors + metformin vs metformin without sodium glucose co-transporter 2 inhibitors (55 vs 115 deaths; 0.65% vs 1.15% per year; HR: 0.52 [95% CI: 0.38-0.72]). No significant differences were observed for major cardiovascular events. Progression to chronic kidney disease stage 4 to 5 was borderline higher with sodium glucose co-transporter 2 inhibitors (HR: 1.28 [95% CI: 0.99-1.64]). ASCVD = atherosclerotic cardiovascular disease; CKD = chronic kidney disease; MI = myocardial infarction; SGLT2i = sodium glucose co-transporter 2 inhibitor; VF = ventricular fibrillation; VT = ventricular tachycardia. Baseline Characteristics of Adults With Endometriosis and Type 2 Diabetes Treated With SGLT2 Inhibitors or Metformin, Before and After Propensity Score Matching ACE = angiotensin-converting enzyme; ARNI = angiotensin receptor–neprilysin inhibitor; BP = blood pressure; COPD = chronic obstructive pulmonary disease; DPP4 = dipeptidyl peptidase-4 inhibitors; GFR = glomerular filtration rate; GLP-1, glucagon-like peptide-1; HbA1c = hemoglobin A1c; HDL = high-density lipoprotein; LDL = low-density lipoprotein; MDRD = Modification of Diet in Renal Disease; MRA = mineralocorticoid receptor antagonist; SGLT2i = sodium glucose co-transporter 2 inhibitor; Std diff. = standardized difference. SGLT2is + Metformin vs Metformin Cardiovascular outcomes in endometriosis patients endometriosis affects ∼10% of reproductive-age women (∼190 million worldwide) and is characterized by ectopic proliferation of endometrial-like tissue outside the uterine cavity. Emerging data suggest increased atherosclerotic cardiovascular disease risk in affected women. In this retrospective, propensity score–matched cohort study using the TriNetX federated research network, adults with endometriosis and type 2 diabetes were identified and classified by treatment with sodium glucose co-transporter 2 inhibitors plus background metformin vs metformin-based therapy without sodium glucose co-transporter 2 inhibitors. After matching, 6.144 patients (3.072 per group) were followed for 2.8 ± 1.8 years, median 2.8, IQR: 3.7. All-cause mortality was significantly lower with sodium glucose co-transporter 2 inhibitors + metformin vs metformin without sodium glucose co-transporter 2 inhibitors (55 vs 115 deaths; 0.65% vs 1.15% per year; HR: 0.52 [95% CI: 0.38-0.72]). No significant differences were observed for major cardiovascular events. Progression to chronic kidney disease stage 4 to 5 was borderline higher with sodium glucose co-transporter 2 inhibitors (HR: 1.28 [95% CI: 0.99-1.64]). ASCVD = atherosclerotic cardiovascular disease; CKD = chronic kidney disease; MI = myocardial infarction; SGLT2i = sodium glucose co-transporter 2 inhibitor; VF = ventricular fibrillation; VT = ventricular tachycardia. After propensity score matching, 3,072 participants were included in each treatment group with excellent covariate balance across demographics, risk factors, comorbidities, laboratory values, and concomitant medications (all standardized differences ≤8%) ( Table 2 ). Table 2 Clinical Outcomes in the Propensity Score–Matched Cohort During Follow-Up (2.8 ± 1.8 Years, Median 2.8, IQR: 3.7) SGLT2i (n = 3,072) Metformin (n = 3,072) HR (95% CI) P Value Number of Events Yearly Rate, % Number of Events Yearly Rate, % Death 55 0.65 115 1.15 0.524 (0.380-0.723) <0.0001 Ischemic stroke or thromboembolism 53 0.69 57 0.66 1.033 (0.711-1.503) 0.86 Acute MI 27 0.29 26 0.28 1.141 (0.665-1.958) 0.63 Incident AF 48 0.58 66 0.77 0.811 (0.559-1.177) 0.27 VT/VF/cardiac arrest 44 0.46 45 0.45 1.06 (0.699-1.608) 0.78 Incident HF 81 0.98 108 1.26 0.818 (0.613-1.092) 0.17 Acute pulmonary edema or cardiogenic shock 27 0.28 22 0.21 1.307 (0.744-2.298) 0.35 Incident PAD 129 1.68 168 2.08 0.855 (0.679-1.075) 0.18 CKD stage 4-5 134 1.53 116 1.34 1.276 (0.995-1.637) 0.05 Incident cancer 147 5.80 171 5.43 0.992 (0.795-1.236) 0.94 AF = atrial fibrillation; CKD = chronic kidney disease; HF = heart failure; MI = myocardial infarction; PAD = peripheral artery disease; VF = ventricular fibrillation; VT = ventricular tachycardia; other abbreviation as in Table 1 . Clinical Outcomes in the Propensity Score–Matched Cohort During Follow-Up (2.8 ± 1.8 Years, Median 2.8, IQR: 3.7) AF = atrial fibrillation; CKD = chronic kidney disease; HF = heart failure; MI = myocardial infarction; PAD = peripheral artery disease; VF = ventricular fibrillation; VT = ventricular tachycardia; other abbreviation as in Table 1 . During a mean follow-up of 2.8 (±1.8) years (median, 2.8; IQR: 3.7), SGLT2i therapy added to background metformin was associated with a lower risk of all-cause mortality compared with metformin-based therapy without SGLT2i (55 vs 115 deaths; annualized rate, 0.65% vs 1.15%; HR: 0.52; 95% CI: 0.38-0.72; P < 0.0001). The multivariable-adjusted analysis for all-cause death yielded results consistent with the propensity-matched analysis (unmatched multivariable HR 0.433 [0.321-0.585], P < 0.0001 vs matched HR 0.524 [0.380-0.723], P < 0.0001), supporting the general robustness of the analyses or associations. Cardiovascular risks were otherwise similar between groups for ischemic stroke or thromboembolism (HR: 1.03; 95% CI: 0.71-1.50; P = 0.86), acute myocardial infarction (HR: 1.14; 95% CI: 0.67-1.96; P = 0.63), ventricular tachycardia or ventricular fibrillation/cardiac arrest (HR: 1.06; 95% CI: 0.70-1.61; P = 0.78), incident HF (HR: 0.82; 95% CI: 0.61-1.09; P = 0.17), acute pulmonary edema or cardiogenic shock (HR: 1.31; 95% CI: 0.74-2.30; P = 0.35), and incident peripheral artery disease (HR: 0.86; 95% CI: 0.68-1.08; P = 0.18). Progression to CKD stage 4 to 5 was borderline higher with SGLT2i (HR: 1.28; 95% CI: 1.00-1.64; P = 0.05).

Large, randomized trials and meta-analyses show consistent reductions in HF events and cardiovascular death with SGLT2i, with benefits observed irrespective of diabetes status or kidney function. Our endometriosis-focused analysis extends these observations to a population underrepresented in trials. This is of public health relevance, as endometriosis affects approximately 10% of women of reproductive age; an estimated 190 million worldwide. In this real-world, propensity score–matched cohort of adults with endometriosis and T2D, SGLT2i use was associated with a 48% lower relative risk of all-cause death compared with metformin over a median of 2.8 years, with no significant differences in major atherosclerotic cardiovascular events. The mortality finding is concordant with trial-level syntheses demonstrating reduced cardiovascular death with SGLT2i across diverse populations, 14 , 15 whereas the absence of a signal for atherosclerotic events aligns with meta-analytic estimates. 13 The mortality reduction appeared early during follow-up despite substantial baseline cardiometabolic risk and obesity, mirroring the rapid onset of benefit reported in randomized trials and meta-analyses across patients with T2D, CKD, and HF, irrespective of baseline cardiovascular risk. 13 , 16 Two findings warrant comment. First, there was no statistically significant association with incident HF. This null result contrasts with trial evidence of HF benefit with SGLT2i and may reflect limited power, modest event rates, and residual confounding despite matching. Second, progression to CKD stage 4 to 5 was nominally higher among SGLT2i users, a counterintuitive finding given the class’s established renoprotective effect. This signal should not be interpreted as evidence of harm. Before matching, baseline kidney disease was substantially more prevalent in the SGLT2i cohort than in the metformin cohort, consistent with confounding by indication, whereby patients with greater underlying renal vulnerability and cardiometabolic complexity were preferentially treated with SGLT2i. Although matching balanced measured covariates, residual confounding likely persists, including unmeasured renal disease severity, albuminuria, longitudinal estimated glomerular filtration rate trajectories, differential surveillance, and competing risks. Collectively, these considerations argue against a causal interpretation of the observed association with CKD stage 4 to 5 progression. Because endometriosis is characterized by low-grade inflammation, oxidative stress, and potentially endothelial dysfunction, features that may amplify cardiometabolic risk and influence responses to glucose-lowering therapies. The extent to which the pleiotropic effects of SGLT2i confer clinical benefit in patients with endometriosis, with or without T2D, has not yet been established. Although endometriosis has been linked to higher risk of atherosclerotic cardiovascular disease, adding this diagnosis to standard risk estimators [eg, Pooled Cohort Equations, SCORE2 (Systematic COronary Risk Evaluation 2)] is unlikely to materially improve prediction in midlife, when traditional factors such as hypertension, dyslipidemia, obesity, and glycemic status predominate. Accordingly, the clinical priority is less recalibration of risk tools and more optimization of therapy selection and timing. In women with endometriosis who develop T2D or exhibit cardiorenal risk features (eg, hypertension, albuminuria, reduced estimated glomerular filtration rate, or a history of HF), earlier initiation of SGLT2i may offer dual benefits for glycemic management and cardiorenal protection. Prospective studies are needed to define thresholds for initiation, quantify absolute risk reductions, and evaluate safety and tolerability in this population, including considerations related to reproductive planning and potential interactions with hormonal therapies. 17 Pragmatic trials that enroll women with endometriosis and elevated cardiometabolic risk, ideally with biomarker enrichment for inflammation or endothelial dysfunction, could determine whether SGLT2i provide benefit beyond standard care centered on metformin and risk factor modification. Finally, the diagnostic landscape for endometriosis is changing. Historically, the mean diagnostic delay has been approximately 7 years, contributing to under-recognition and misclassification in observational cohorts and limiting timely cardiometabolic risk mitigation. A saliva-based microRNA signature, that is already available in more than 20 countries with results available within a few days, demonstrates high diagnostic performance (sensitivity, 97.3% [95% CI: 96.4%-98.0%]; specificity, 94.1% [95% CI: 91.0%-96.4%]) and substantially lower misclassification than imaging alone among surgically confirmed cases (4.6% vs 27.2%). 18 This shift has direct implications for cardiometabolic research and care. Earlier and more accurate identification of endometriosis will also facilitate improved characterization and optimization of hormonal treatment strategies (including gonadotropin-releasing hormone agonists, combined oral contraceptives, and progestins), which are known to improve quality of life and reduce symptoms but also influence cardiovascular risk profiles and may interact with glycemic control and glucose-lowering therapies. In parallel, broader and earlier identification of endometriosis will facilitate proactive outreach to women with T2D for risk assessment, optimization of preventive therapies, and enrollment in comparative effectiveness studies. This observational analysis has several important limitations, particularly in the context of the robust randomized evidence supporting the cardiovascular and renal benefits of SGLT2i. Confounding, misclassification, and selection bias are inherent to nonrandomized study designs. Electronic health record–derived phenotypes may incompletely capture disease severity; particularly for endometriosis, for which grading, anatomic localization, self-reported diagnoses, and imaging extent are not recorded; as well as specific medication adherence and dosing. In addition, key treatment-related variables, including medication adherence, dosing, treatment persistence, and temporal changes in therapy, are not captured. Although propensity score matching substantially improved balance across measured covariates, this approach necessarily resulted in the exclusion of a substantial proportion of SGLT2i-treated patients and limits external validity. The matched cohort therefore represents a subset of SGLT2i users who were most comparable to metformin users with respect to measured characteristics, and the findings should not be extrapolated uncritically to all SGLT2i-treated patients in the source population. Moreover, residual and unmeasured confounding related to disease severity, treatment indication, duration of diabetes, and baseline cardiovascular and renal risk likely persists, reflecting indication bias at treatment initiation that cannot be fully eliminated in registry-based analyses. Use of the TriNetX research network relies on routine clinical coding and site-level data capture, which may result in heterogeneous laboratory ascertainment and variable data completeness across centers. These limitations are particularly relevant for renal outcomes, for which modest declines in kidney function, longitudinal albuminuria trajectories, and competing risks may not be fully captured. Finally, the absence of statistically significant associations for several outcomes may reflect limited statistical power, low absolute event rates after matching, and the highly selected nature of the study population, which consisted predominantly of White, middle-aged individuals with a high prevalence of obesity. Accordingly, these findings should be interpreted as hypothesis-generating and not as contradictory to evidence from randomized outcome trials.

Among patients with endometriosis and T2D, use of SGLT2i in combination with metformin was associated with lower mortality, while rates of major atherosclerotic cardiovascular events were similar to metformin-based therapy alone. Perspectives COMPETENCY IN MEDICAL KNOWLEDGE: In a large, propensity score–matched cohort of patients with endometriosis and T2D, treatment with SGLT2i added to background metformin was associated with substantially lower all-cause mortality compared with metformin-based therapy without SGLT2 inhibition, while rates of MACE and progression to advanced CKD did not differ significantly between groups. These findings suggest that SGLT2i may offer a mortality advantage in this population. TRANSLATIONAL OUTLOOK: Future investigations should elucidate the mechanisms driving the observed survival benefit and define whether—and through which pathways—SGLT2 inhibition modifies endometriosis-related inflammation or metabolic dysregulation. Basic and translational studies focused on the interaction between SGLT2 inhibition added to standard glucose-lowering therapy and endometriosis-associated cardiometabolic risk, along with prospective clinical trials, are needed to refine therapeutic strategies and resolve remaining uncertainties in this population. COMPETENCY IN MEDICAL KNOWLEDGE: In a large, propensity score–matched cohort of patients with endometriosis and T2D, treatment with SGLT2i added to background metformin was associated with substantially lower all-cause mortality compared with metformin-based therapy without SGLT2 inhibition, while rates of MACE and progression to advanced CKD did not differ significantly between groups. These findings suggest that SGLT2i may offer a mortality advantage in this population. TRANSLATIONAL OUTLOOK: Future investigations should elucidate the mechanisms driving the observed survival benefit and define whether—and through which pathways—SGLT2 inhibition modifies endometriosis-related inflammation or metabolic dysregulation. Basic and translational studies focused on the interaction between SGLT2 inhibition added to standard glucose-lowering therapy and endometriosis-associated cardiometabolic risk, along with prospective clinical trials, are needed to refine therapeutic strategies and resolve remaining uncertainties in this population.

This research was supported by GERCA (Groupe pour l’Enseignement et la Recherche Cardiovasculaire en Alsace, Strasbourg, France), a charitable cardiovascular research consortium. Dr Morel has received research grants supporting investigator-initiated studies from AstraZeneca, Medtronic, and Boehringer Ingelheim, all outside the submitted work; and has also received funding from Fondation Coeur et Recherche and EndoFrance, two recognized French charities dedicated to advancing research on cardiovascular disease and endometriosis. Dr Fauchier has received modest consultancy or speaker fees from AstraZeneca, Bayer, BMS/Pfizer, Boehringer Ingelheim, Boston Scientific, Medtronic, Novo Nordisk, and Zoll, all outside the submitted work. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.