Blood Clotting Abnormalities Linked to Poorer IVF Outcomes in Endometriosis

Infertile women with surgically confirmed endometriosis showed markedly disturbed blood coagulation profiles before starting in vitro fertilization, and two of those disturbances—an elevated D-dimer level and reduced antithrombin III activity—were independently associated with failure to achieve a live birth, according to a retrospective cohort study conducted at Shijiazhuang Maternal and Child Health Care Hospital in Hebei, China. Using a composite definition based on fibrinogen, D-dimer, and antithrombin III, the researchers found that 48.15 percent of 108 endometriosis patients met criteria for at least one coagulation abnormality, compared with just 7.41 percent of 108 age-matched controls with tubal-factor infertility, a difference the authors reported as statistically significant. The study, published in Frontiers in Medicine, is among the first to systematically connect a panel of coagulation markers with embryo quality, implantation, and live birth in this population, although the authors emphasize that the findings are exploratory and do not yet justify changes to clinical practice.

Endometriosis is an estrogen-dependent chronic disease in which tissue resembling the uterine lining grows outside the uterine cavity, affecting an estimated 6 to 10 percent of women of childbearing age. Its overlap with infertility is striking: the prevalence of endometriosis among infertile patients reaches 30 to 50 percent, and 40 to 50 percent of women with the disease experience difficulty conceiving. In vitro fertilization with embryo transfer is the first-line assisted reproductive technology for endometriosis-related infertility, yet clinical observations have long shown that these patients tend to have lower implantation rates, higher early pregnancy loss, and reduced live birth rates compared with women whose infertility stems from other causes, such as damaged fallopian tubes. The mechanisms behind this gap remain incompletely understood, and the study’s authors argue that coagulation dysfunction deserves closer scrutiny as a possible contributor.

The rationale rests on an evolving view of endometriosis as more than a localized pelvic condition. Accumulating evidence characterizes it as a systemic chronic inflammatory disease, and persistent low-grade inflammation can activate the coagulation cascade, suppress fibrinolysis, and push the body toward a prothrombotic state. Coagulation dysfunction may in turn impair the microcirculation of the ovary and endometrium, potentially compromising follicular development and the receptivity of the uterine lining, and it may raise the risk of early miscarriage after an embryo implants. A recent systematic review and meta-analysis of 22 studies including 6,258 participants confirmed that prothrombin time, activated partial thromboplastin time, and thrombin time are significantly shortened in endometriosis compared with controls, supporting the biological plausibility of a hypercoagulable phenotype in these patients.

To test whether coagulation status tracks with IVF outcomes, the team screened 241 patients between January 2022 and December 2024 and ultimately analyzed 216 women who completed controlled ovarian hyperstimulation and underwent embryo transfer: 108 with laparoscopically confirmed endometriosis and 108 age-matched controls with tubal-factor infertility only. Every endometriosis diagnosis was made surgically, with disease staged by the revised American Fertility Society classification from operative records; 42 patients had stage I–II disease and 66 had stage III–IV disease. To ensure statistical independence, only each patient’s first IVF cycle and first embryo transfer were included. The groups were well balanced on age, body mass index, infertility duration, and ovarian reserve markers including basal follicle-stimulating hormone, anti-Mullerian hormone, and antral follicle count, and treatment-related factors such as stimulation protocol, fertilization method, transfer stage, and transfer type were comparable.

Fasting venous blood was drawn on days 2 to 3 of the menstrual cycle, before ovarian stimulation began, and analyzed on an automated Sysmex CS-5100 coagulation analyzer. The panel covered prothrombin time, activated partial thromboplastin time, thrombin time, fibrinogen, D-dimer, and antithrombin III activity. The researchers defined coagulation abnormality as at least one of three findings: fibrinogen above 4.0 grams per liter, D-dimer above 0.5 milligrams per liter, or antithrombin III below 75 percent. Notably, the three clotting times were measured and analyzed as continuous variables but were deliberately excluded from the classification, because no internationally standardized cutoff exists for defining shortened values as a hypercoagulable state. The composite definition was chosen to maximize sensitivity for detecting a prothrombotic tendency, on the reasoning that coagulation activation in endometriosis is likely a continuum rather than a binary state.

The coagulation data confirmed a clear biochemical divide between the groups. Endometriosis patients had significantly shortened prothrombin time, activated partial thromboplastin time, and thrombin time, elevated fibrinogen and D-dimer, and reduced antithrombin III activity compared with controls, all statistically significant. Mean D-dimer in the endometriosis group was 1.92 milligrams per liter against 0.36 in controls, while mean antithrombin III activity was 68.52 percent versus 96.35 percent. When patients were stratified by the composite definition, the endometriosis group split into 52 women with coagulation abnormalities and 56 with normal coagulation, and the two subgroups were similar in age, ovarian reserve, and stimulation characteristics, making the outcome contrasts more interpretable.

Those contrasts were substantial. Women with endometriosis and coagulation abnormalities had a high-quality embryo rate of 43.1 percent versus 54.7 percent among those with normal coagulation, an implantation rate of 26.2 percent versus 39.7 percent, and a live birth rate of 21.2 percent versus 44.6 percent. Early abortion rates ran in the opposite direction, at 42.1 percent of clinical pregnancies in the abnormal-coagulation subgroup versus 16.1 percent. Fertilization and cleavage rates, however, did not differ significantly between subgroups, a pattern the authors interpret as pointing away from impaired gamete interaction or early embryonic cell division and toward damage further downstream, possibly involving endometrial receptivity and early placentation. Live birth rates also declined progressively with the number of abnormal parameters among fibrinogen, D-dimer, and antithrombin III, falling from 44.6 percent with none abnormal to 29.6 percent with one, 16.7 percent with two, and 0.0 percent with all three, a trend the authors reported as statistically significant.

Hierarchical logistic regression sharpened the picture. In a model adjusting for age, body mass index, endometriosis stage, stimulation protocol, transfer stage, and transfer type, elevated D-dimer was associated with live birth failure with an odds ratio of 3.412, and decreased antithrombin III with an odds ratio of 0.452. Adding high-quality embryo rate to the model attenuated the D-dimer estimate by about 16 percent, to an odds ratio of 2.864, but the association remained statistically significant, as did antithrombin III at an odds ratio of 0.512, while high-quality embryo rate itself was protective. The authors stress that this attenuation reflects a change in the adjusted association rather than evidence of causality. Fibrinogen was not independently significant. Sensitivity analyses reinforced the findings: when the markers were treated as continuous variables, each 1 milligram per liter increase in D-dimer carried an odds ratio of 1.721 for live birth failure, and each 1 percent increase in antithrombin III an odds ratio of 0.963. Under a stricter definition requiring at least two concurrent abnormalities, 25 patients were affected, their live birth rate fell to 12.0 percent versus 39.7 percent, and the adjusted odds ratio reached 3.116.

Subgroup analysis hinted that the association was stronger in advanced disease, with an odds ratio of 3.214 for live birth failure among stage III–IV patients compared with 1.876 in stage I–II, but the formal interaction test was not statistically significant, so the authors treat this as hypothesis-generating only. The study carries important limitations. It was single-center and retrospective, with a modest sample size and potential selection bias; residual confounding cannot be excluded, particularly for unmeasured factors such as embryonic chromosomal abnormalities or uterine structural problems. Laparoscopy was not performed in controls, so subclinical endometriosis could have been misclassified, which would tend to blunt rather than exaggerate the between-group differences. Excluding the few patients who developed ovarian hyperstimulation syndrome or had cycle cancellations may also have selected women who progressed to transfer. The authors note that the broad primary definition pools mild and severe derangements, and that the dose-response gradient, while consistent, requires confirmation.

The clinical message is deliberately restrained. Current European Society of Human Reproduction and Embryology guidelines do not recommend routine coagulation testing before IVF in women with endometriosis, and the authors state plainly that routine D-dimer and antithrombin III testing cannot be recommended on the basis of this study. Because the design was observational, the regression models cannot establish that coagulation abnormalities cause IVF failure, and any anticoagulation strategy would need to be weighed against bleeding risks in properly controlled trials. What the study does provide is a detailed, quantified map of how frequently coagulation disturbances occur in this population and how strongly they track with live birth outcomes, offering preliminary evidence for future research on perioperative risk stratification. The authors call for multicenter prospective studies with larger samples to validate the composite definition, confirm the dose-response pattern, and determine whether targeted intervention can translate these associations into improved reproductive outcomes.

Subject of Research: Coagulation abnormalities and their association with in vitro fertilization outcomes in infertile patients with endometriosis

Article Title: Correlation between coagulation abnormalities and in vitro fertilization outcomes in infertile patients with endometriosis: a retrospective cohort study

Article References: Du, X., Hu, W., & Fu, C. (2026). Correlation between coagulation abnormalities and in vitro fertilization outcomes in infertile patients with endometriosis: a retrospective cohort study. Frontiers in Medicine, 13, Article 1888342. https://doi.org/10.3389/fmed.2026.1888342

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1888342

Keywords: endometriosis, in vitro fertilization, coagulation, D-dimer, antithrombin III, fibrinogen, infertility, live birth rate, embryo quality, retrospective cohort study, hypercoagulable state, reproductive medicine