Abstract
Objective
This study aims to investigate the association between the dynamics of routine metabolic markers and endometriosis severity.
Methods
A retrospective analysis was conducted on patients diagnosed with endometriosis at Zhongshan Hospital, Xiamen, affiliated with Fudan University. The collected data encompassed demographic details and biochemical indicators related to lipid, hepatobiliary, renal metabolism, and electrolyte balance. Independent influencing factors were screened by univariate logistic regression and statistically significant variables were included in the model for adjustment. Restricted cubic spline (RCS) models were also plotted to analyze the nonlinear relationship between factors and endometriosis severity. The receiver operating characteristic (ROC) curve was used to validate the discriminative ability of independent influencing factors.
Results
Ninety-four patients were enrolled in the study, including 32 at stage IV as classified by the American Society for Reproductive Medicine (ASRM) staging. Univariate analysis identified fasting blood glucose (FBG), total protein, direct bilirubin, total bilirubin (TBil) and glutamic-pyruvic transaminase (ALT) as significant metabolic indicators. Additionally, carbohydrate antigen 125 (CA125) and human epididymal protein 4 (HE4) emerged as significant covariates. The RCS analysis revealed a nonlinear association between most metabolic indicators and outcome measures. ROC curve analysis showed that the area under the curve (AUC) of the alanine transaminase (ALT) was above 0.6.
Conclusion
ALT had a negative correlation with the severity of endometriosis and was an independent influencing factor with statistical significance. This finding could offer clinicians non-invasive biomarkers for early detection and precise monitoring of disease progression.
Keywords: endometriosis, metabolic, American Society for Reproductive Medicine staging
Introduction
Endometriosis, a chronic inflammatory disorder characterized by the presence of endometrial-like tissue ectopic to the uterus, is linked to pelvic pain and infertility.1,2 It is estrogen-dependent, prevalent during the reproductive years, and affects 5–15% of women globally.3,4 This disorder poses a serious public health problem and economic strain.5 According to Ballweg et al, the average wait time for a final diagnosis of endometriosis is nine years.6 There is growing evidence that endometriosis raises the risk of several pregnancy-related complications, including premature placental abruption, retained placenta, premature rupture of membranes, pre-eclampsia, pregnancy-induced hypertension, gestational diabetes mellitus, gestational cholestasis, antepartum and postpartum hemorrhages, labor dystocia, stillbirth, neonatal deaths, and uterine congenital abnormalities.7 In a few cases, endometriosis can undergo malignant transformation, with ovarian cancer being the most frequent malignancy associated with the disease.8–10
Deep endometriotic lesions have the potential to infiltrate nerves11 and lymph nodes,12 causing heightened negative effects on the body. However, research on the severity of endometriosis is currently limited. Existing studies have identified advancing age,13 concomitant autoimmune diseases, and the frequency of laparoscopic operations as predictors of endometriosis severity.14 To date, a few studies have explored the correlation between patients’ metabolic profiles and the severity of their endometriosis. The liver plays a central role in regulating systemic metabolism, including glucose homeostasis. Dysregulation in liver function can lead to imbalances in metabolic pathways that influence inflammation and immune responses, both of which are implicated in the development and progression of endometriosis.15 Additionally, altered glucose metabolism can affect energy availability and cellular function in endometrial tissue, potentially contributing to the survival and growth of ectopic endometrial implants.16 Thus, understanding these metabolic connections may provide insights into the underlying mechanisms of endometriosis.
In this study, we aimed to investigate the correlation between metabolic indicators and the severity of endometriosis using both univariate and multivariate logistic regression analyses. Additionally, restricted cubic spline modeling was applied to examine nonlinear relationships. This research may provide valuable early diagnostic markers and therapeutic strategies for severe endometriosis.
Materials and Methods
Research Cohort and Profile
This study retrospectively collected patients diagnosed with endometriosis by laparoscopy or laparotomy based on histological confirmation in Zhongshan Hospital (Xiamen), Fudan University from January 2018 to August 2022. Patients were excluded from the study if they presented with abnormal metabolic markers, hypertension, diabetes, hyperlipidemia, liver or gallbladder diseases, autoimmune diseases, a history of uterine surgery or pregnancy, hormone therapy, or if there was any missing information. The collected variables included covariates and indicators reflecting patient lipid metabolism, hepatobiliary metabolism, renal metabolism, and electrolyte metabolism. Covariates included age, body mass index (BMI), carbohydrate antigen 125 (CA-125), human epididymis protein 4 (HE4). Metabolic indicators included apolipoprotein A, apolipoprotein B, fasting blood glucose, serum albumin, serum total protein, direct bilirubin, total bilirubin, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, γ-glutamyl transferase, lactate dehydrogenase, prealbumin, urea, creatinine, glomerular filtration rate, uric acid, sodium, potassium, chloride, CO2, total cholesterol, triglycerides, HDL, and LDL. All laboratory data were collected within 3 days of the end of the patient’s menstrual period. ASRM staging data for endometriosis were collected from patients, with all diagnoses confirmed through pathological examination. This study was performed in accordance with the declaration of Helsinki and was approved by the ethics committee of Xiamen Hospital, Zhongshan Hospital, Fudan University.
Statistical Analysis
Categorical variables were described using frequency and percentage (%), with the chi-square test was used to compare the differences between the two groups. Continuous variables were tested for normality. Continuous variables with normal distribution were described using mean and standard deviation (Mean (SD)), and group differences were compared using a t-test. However, non-normally distributed continuous variables were described using medians and quartiles (Median [IQR]), and the differences between the two groups were compared using the rank sum test. Independent factors influencing endometriosis severity were ascertained by univariate logistic regression. Notably, according to the results of univariate regression and stepwise regression combined with factors that were known or suspected to be related to endometriosis severity, we finally determined the variable selection in multivariate models. Moreover, restricted cubic spline models were developed to analyze the nonlinear relationship between metabolic indicators and outcomes. A nomograph was drawn to visualize the independent influencing factors, and the ROC curve was used to verify the discriminative ability of the independent influencing factors. All statistical analyzes were performed using R 4.2.1 (https://www.r-project.org), and a double trailed P value < 0.05 was considered statistically significant.
Results
Patient Characteristics
In accordance with the patient inclusion criteria, this study included a total of 94 endometriosis patients, 32 of whom were diagnosed with ASRM stage IV. The mean age of all patients was 34.85 years old, with the ASRM stage IV patients having a mean age of 36.81 years. Table 1 offers a comprehensive summary of the demographic and clinical characteristics of the patients.
Table 1.
The Characteristics of All Patients
| Characteristics | ASRM Staging I/II/III | ASRM Staging IV | P-value |
|---|---|---|---|
| n | 62 | 32 | |
| Age (mean (SD)) | 34.85 (7.00) | 36.81 (8.42) | 0.234 |
| CA125 | 48.78 [31.12, 69.22] | 72.94 [35.71, 127.00] | 0.020 |
| HE4 | 47.65 [42.31, 55.56] | 47.72 [38.33, 54.10] | 0.346 |
| Apolipoprotein A (g/L) | 1.32 [1.28, 1.39] | 1.37 [1.29, 1.43] | 0.094 |
| Apolipoprotein B (g/L) | 0.85 [0.74, 0.89] | 0.89 [0.77, 0.98] | 0.140 |
| FBG | 4.88 [4.70, 5.00] | 4.95 [4.77, 5.43] | 0.100 |
| Albumin (g/L) | 45.00 [42.25, 47.00] | 45.00 [43.75, 48.00] | 0.289 |
| Total protein (g/L) | 71.00 [67.25, 74.00] | 72.50 [69.00, 77.00] | 0.094 |
| Direct bilirubin (μmol/L) | 3.40 [2.73, 4.27] | 2.75 [2.48, 3.42] | 0.022 |
| TBil (μmol/L) | 8.40 [6.40, 10.55] | 7.25 [5.75, 8.07] | 0.022 |
| ALT (U/L) | 12.00 [10.00, 16.00] | 11.00 [8.00, 12.00] | 0.049 |
| AST (U/L) | 16.00 [14.00, 18.00] | 14.50 [13.00, 17.00] | 0.145 |
| ALP (U/L) | 53.00 [44.25, 59.00] | 52.50 [45.75, 65.50] | 0.500 |
| γ-GT (U/L) | 13.00 [11.00, 17.00] | 13.00 [11.00, 17.25] | 0.764 |
| LDH (U/L) | 151.50 [138.00, 171.25] | 166.00 [148.75, 175.50] | 0.063 |
| Prealbumin (g/L) | 0.21 [0.19, 0.23] | 0.21 [0.18, 0.23] | 0.758 |
| Urea (mmol/L) | 4.25 [3.23, 4.97] | 4.05 [3.45, 5.03] | 0.734 |
| Creatinine (μmol/L) | 58.40 (8.65) | 58.09 (8.38) | 0.868 |
| GFR (mL/min/1.73 m2) | 112.66 (11.51) | 111.81 (12.80) | 0.746 |
| Uric Acid (μmol/L) | 279.00 [226.00, 320.25] | 266.00 [241.75, 306.25] | 0.914 |
| Sodium (mmol/L) | 140.00 [139.00, 141.00] | 140.00 [139.00, 141.00] | 0.607 |
| Potassium (mmol/L) | 4.00 [3.80, 4.10] | 4.05 [3.90, 4.20] | 0.155 |
| Chlorine (mmol/L) | 104.00 [102.00, 105.00] | 104.00 [102.00, 105.00] | 0.955 |
| CO2 (mmol/L) | 24.84 (1.79) | 24.59 (2.01) | 0.548 |
| Total cholesterol (mmol/L) | 4.37 [3.95, 4.61] | 4.43 [4.01, 4.69] | 0.273 |
| Triglyceride (mmol/L) | 0.99 [0.82, 1.15] | 1.10 [0.85, 1.31] | 0.205 |
| HDL (mmol/L) | 1.47 [1.36, 1.57] | 1.50 [1.44, 1.57] | 0.222 |
| LDL (mmol/L) | 2.40 [2.25, 2.61] | 2.57 [2.21, 2.78] | 0.161 |
| BMI | 21.02 [19.59, 22.51] | 21.39 [19.15, 23.19] | 0.658 |
Notes: The data were presented as mean (standard deviation) or median [interquartile range].
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; γ-GT, γ-glutamyltransferase; LDH, lactate dehydrogenase; GFR, Glomerular filtration rate; HDL, high-density lipoprotein; LDL, Low-density lipoprotein; BMI, body mass index.
Influence of Metabolic Indicators on the Severity of Endometriosis
To analyze the effect of different levels of metabolic indicators on outcome indicators, we categorized metabolic indicators according to their quartiles and included them in logistic regression for analysis in the form of both continuous and categorical variables. The results of the univariate logistic regression showed that FBG (OR [95% CI]: Q4: 3.5[1.093, 11.974], continuous: 3.422[1.116, 11.539]), total protein (OR [95% CI]: continuous: 1.094[1.012, 1.198]), direct bilirubin (OR [95% CI]: Q4: 0.176[0.035, 0.683], continuous: 0.645[0.402, 0.972]), TBil (OR [95% CI]: Q4: 0.278[0.073, 0.933]) and ALT (OR [95% CI]: Q4: 0.239[0.049, 0.888]) were statistically significant in relation to the severity of endometriosis (Table 2).
Table 2.
The Results of the Univariate Analysis
| Characteristics | OR [95% CI] | P-value | ||
|---|---|---|---|---|
| Metabolic markers | Apolipoprotein A (continuous) | 10.696[0.426, 468.785] | 0.1789 | |
| Apolipoprotein A | Q1 | – | – | |
| Q2 | 0.675[0.170, 2.521] | 0.5605 | ||
| Q3 | 1.295[0.377, 4.533] | 0.6800 | ||
| Q4 | 2.429[0.754, 8.311] | 0.1437 | ||
| Apolipoprotein B (continuous) | 5.096[0.413, 82.358] | 0.2145 | ||
| Apolipoprotein B | Q1 | – | – | |
| Q2 | 1.295[0.377, 4.533] | 0.6800 | ||
| Q3 | 0.511[0.117, 2.000] | 0.3448 | ||
| Q4 | 2.870[0.893, 9.876] | 0.0828 | ||
| FBG (continuous) | 3.422[1.116, 11.539] | 0.0372 | ||
| FBG | Q1 | – | – | |
| Q2 | 1.105[0.342, 3.574] | 0.8657 | ||
| Q3 | 0.808[0.185, 3.130] | 0.7625 | ||
| Q4 | 3.500[1.093, 11.974] | 0.0387 | ||
| Albumin (continuous) | 1.083[0.959, 1.233] | 0.2107 | ||
| Albumin | Q1 | – | – | |
| Q2 | 2.352[0.717, 7.974] | 0.1599 | ||
| Q3 | 1.327[0.366, 4.684] | 0.6595 | ||
| Q4 | 1.725[0.544, 5.589] | 0.3541 | ||
| Total protein (continuous) | 1.094[1.012, 1.198] | 0.0399 | ||
| Total protein | Q1 | – | – | |
| Q2 | 1.150[0.324, 3.907] | 0.8237 | ||
| Q3 | 0.863[0.249, 2.817] | 0.8085 | ||
| Q4 | 2.091[0.675, 6.633] | 0.2021 | ||
| Direct bilirubin (continuous) | 0.645[0.402, 0.972] | 0.0492 | ||
| Direct bilirubin | Q1 | – | – | |
| Q2 | 0.571[0.174, 1.804] | 0.3444 | ||
| Q3 | 0.444[0.138, 1.359] | 0.1608 | ||
| Q4 | 0.176[0.035, 0.683] | 0.0189 | ||
| TBil (continuous) | 0.853[0.718, 0.991] | 0.0513 | ||
| TBil | Q1 | – | – | |
| Q2 | 0.769[0.245, 2.374] | 0.6483 | ||
| Q3 | 0.222[0.053, 0.788] | 0.0265 | ||
| Q4 | 0.278[0.073, 0.933] | 0.0453 | ||
| ALT (continuous) | 0.918[0.826, 1.001] | 0.0756 | ||
| ALT | Q1 | – | – | |
| Q2 | 1.163[0.411, 3.298] | 0.7746 | ||
| Q3 | 0.339[0.068, 1.314] | 0.1416 | ||
| Q4 | 0.239[0.049, 0.888] | 0.0467 | ||
| AST (continuous) | 0.941[0.826, 1.056] | 0.3234 | ||
| AST | Q1 | – | – | |
| Q2 | 0.412[0.12, 1.327] | 0.1437 | ||
| Q3 | 0.350[0.104, 1.109] | 0.0800 | ||
| Q4 | 0.462[0.125, 1.585] | 0.2274 | ||
| ALP (continuous) | 1.020[0.988, 1.055] | 0.2284 | ||
| ALP | Q1 | – | – | |
| Q2 | 1.195[0.369, 3.928] | 0.7654 | ||
| Q3 | 0.664[0.169, 2.422] | 0.5400 | ||
| Q4 | 1.635[0.506, 5.438] | 0.4133 | ||
| γ-GT (continuous) | 1.004[0.961, 1.045] | 0.8347 | ||
| γ-GT | Q1 | – | – | |
| Q2 | 2.100[0.593, 7.503] | 0.2463 | ||
| Q3 | 0.988[0.304, 3.103] | 0.9839 | ||
| Q4 | 1.477[0.462, 4.695] | 0.5059 | ||
| LDH (continuous) | 1.013[0.997, 1.029] | 0.1085 | ||
| LDH | Q1 | – | – | |
| Q2 | 0.463[0.106, 1.796] | 0.2753 | ||
| Q3 | 2.870[0.893, 9.876] | 0.0828 | ||
| Q4 | 1.495[0.429, 5.324] | 0.5271 | ||
| Prealbumin (continuous) | 5.877[0, 361,026.273] | 0.7498 | ||
| Prealbumin | Q1 | – | – | |
| Q2 | 0.844[0.268, 2.581] | 0.7677 | ||
| Q3 | 0.565[0.151, 1.913] | 0.3718 | ||
| Q4 | 0.792[0.240, 2.501] | 0.6930 | ||
| Urea (continuous) | 1.051[0.702, 1.575] | 0.8078 | ||
| Urea | Q1 | – | – | |
| Q2 | 1.558[0.474, 5.232] | 0.4654 | ||
| Q3 | 0.875[0.256, 2.941] | 0.8283 | ||
| Q4 | 1.385[0.409, 4.732] | 0.5988 | ||
| Creatinine (continuous) | 0.996[0.946, 1.047] | 0.8667 | ||
| Creatinine | Q1 | – | – | |
| Q2 | 1.446[0.443, 4.810] | 0.5402 | ||
| Q3 | 1.385[0.409, 4.732] | 0.5988 | ||
| Q4 | 0.926[0.270, 3.132] | 0.9017 | ||
| GFR (continuous) | 0.994[0.959, 1.031] | 0.7428 | ||
| GFR | Q1 | – | – | |
| Q2 | 1.071[0.327, 3.514] | 0.9085 | ||
| Q3 | 0.889[0.266, 2.941] | 0.8463 | ||
| Q4 | 0.556[0.154, 1.895] | 0.3527 | ||
| Uric Acid (continuous) | 1.001[0.994, 1.008] | 0.7581 | ||
| Uric Acid | Q1 | – | – | |
| Q2 | 2.226[0.680, 7.692] | 0.1918 | ||
| Q3 | 1.062[0.300, 3.771] | 0.9243 | ||
| Q4 | 1.000[0.284, 3.527] | 1.0000 | ||
| Sodium (continuous) | 1.056[0.834, 1.344] | 0.6520 | ||
| Sodium | Q1 | – | – | |
| Q2 | 0.481[0.135, 1.512] | 0.2266 | ||
| Q3 | 3.846[1.015, 16.755] | 0.0548 | ||
| Q4 | 0.888[0.261, 2.828] | 0.8426 | ||
| Potassium (continuous) | 3.534[0.916, 16.839] | 0.0897 | ||
| Potassium | Q1 | – | – | |
| Q2 | 2.787[0.823, 10.511] | 0.1096 | ||
| Q3 | 2.235[0.656, 8.419] | 0.2098 | ||
| Q4 | 2.073[0.511, 8.799] | 0.3075 | ||
| Chlorine (continuous) | 0.959[0.777, 1.183] | 0.6963 | ||
| Chlorine | Q1 | – | – | |
| Q2 | 0.871[0.304, 2.505] | 0.7957 | ||
| Q3 | 1.330[0.381, 4.594] | 0.6502 | ||
| Q4 | 0.844[0.190, 3.340] | 0.8135 | ||
| CO2 (continuous) | 0.931[0.735, 1.173] | 0.5435 | ||
| CO2 | Q1 | – | – | |
| Q2 | 0.688[0.233, 2.021] | 0.4934 | ||
| Q3 | 0.682[0.171, 2.517] | 0.5713 | ||
| Q4 | 0.750[0.204, 2.631] | 0.6558 | ||
| Total cholesterol (continuous) | 1.846[0.833, 4.391] | 0.1425 | ||
| Total cholesterol | Q1 | – | – | |
| Q2 | 0.992[0.284, 3.415] | 0.9894 | ||
| Q3 | 0.750[0.206, 2.619] | 0.6530 | ||
| Q4 | 1.798[0.567, 5.911] | 0.3225 | ||
| Triglyceride (continuous) | 1.627[0.567, 4.815] | 0.3578 | ||
| Triglyceride | Q1 | – | – | |
| Q2 | 0.421[0.097, 1.598] | 0.2166 | ||
| Q3 | 1.286[0.389, 4.317] | 0.6795 | ||
| Q4 | 1.692[0.530, 5.592] | 0.3775 | ||
| HDL (continuous) | 2.115[0.270, 20.622] | 0.4862 | ||
| HDL | Q1 | – | – | |
| Q2 | 2.027[0.559, 7.943] | 0.2893 | ||
| Q3 | 3.483[1.006, 13.482] | 0.0561 | ||
| Q4 | 1.900[0.527, 7.400] | 0.3333 | ||
| LDL (continuous) | 2.291[0.869, 6.816] | 0.1092 | ||
| LDL | Q1 | – | – | |
| Q2 | 0.706[0.193, 2.478] | 0.5878 | ||
| Q3 | 0.706[0.193, 2.478] | 0.5878 | ||
| Q4 | 2.000[0.631, 6.622] | 0.2441 | ||
| Covariates | Age (continuous) | 1.035[0.978, 1.098] | 0.2330 | |
| Age | Q1 | – | – | |
| Q2 | 1.490[0.431, 5.372] | 0.5298 | ||
| Q3 | 1.705[0.464, 6.460] | 0.4206 | ||
| Q4 | 2.679[0.812, 9.556] | 0.1132 | ||
| CA125 (continuous) | 1.009[1.003, 1.018] | 0.0167 | ||
| CA125 | Q1 | – | – | |
| Q2 | 0.675[0.170, 2.521] | 0.5605 | ||
| Q3 | 1.062[0.300, 3.771] | 0.9243 | ||
| Q4 | 2.870[0.893, 9.876] | 0.0828 | ||
| HE4 (continuous) | 0.987[0.947, 1.026] | 0.5174 | ||
| HE4 | Q1 | – | – | |
| Q2 | 0.211[0.049, 0.758] | 0.0229 | ||
| Q3 | 0.533[0.160, 1.705] | 0.2937 | ||
| Q4 | 0.500[0.151, 1.586] | 0.2441 | ||
| BMI (continuous) | 1.053[0.908, 1.223] | 0.4888 | ||
| BMI | Q1 | – | – | |
| Q2 | 0.588[0.162, 2.018] | 0.4034 | ||
| Q3 | 0.729[0.211, 2.448] | 0.6099 | ||
| Q4 | 1.190[0.372, 3.848] | 0.7680 | ||
Notes: Q1, 0%–25% quantile; Q2, 25%–50% quantile; Q3, 50%–75% quantile; Q4, 75%–100% quantile. Bolded values indicate statistical significance.
For the above variables, we included covariates for adjustment (Model 1: unadjusted; Model 2: adjusted for age, BMI; Model 3: adjusted for age, BMI, CA125, HE4). The results showed that FBG and total protein were not statistically significant associated with endometriosis severity after adjustment for age and BMI. However, TBil (OR [95% CI]: 0.28[0.073, 0.957], P: 0.0499) and direct bilirubin (OR [95% CI]: 0.18[0.035, 0.702], P: 0.0209) remained significantly associated with endometriosis severity after adjustment for age and BMI. Additionally, ALT (Model 2: OR [95% CI]: 0.194[0.037, 0.768], P: 0.03, Model 3: OR [95% CI]: 0.138[0.019, 0.67], P: 0.0247) remained significantly associated with endometriosis severity after adjustment for age, BMI, CA125, and HE4 (Table 3).
Table 3.
Impact of Metabolic Indicators on Outcome
| Levels | Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|---|
| OR [95% CI] | P-value | OR [95% CI] | P-value | OR [95% CI] | P-value | ||
| TBil (continuous) | 0.853 [0.718, 0.991] | 0.051 | 0.860 [0.723, 1.000] | 0.065 | 0.883 [0.740, 1.030] | 0.130 | |
| P of trend: 0.013* | P of trend: 0.012* | P of trend: 0.016* | |||||
| TBil (categorical) | Q1 | – | – | – | – | – | – |
| Q2 | 0.769[0.245, 2.374] | 0.6483 | 0.739[0.231, 2.317] | 0.6048 | 0.823[0.235, 2.827] | 0.7574 | |
| Q3 | 0.222[0.053, 0.788] | 0.0265* | 0.220[0.05, 0.817] | 0.0308* | 0.281[0.058, 1.155] | 0.0898 | |
| Q4 | 0.278[0.073, 0.933] | 0.0453* | 0.280[0.073, 0.957] | 0.0499* | 0.334[0.08, 1.251] | 0.1131 | |
| Direct bilirubin (continuous) | 0.645[0.402, 0.972] | 0.049* | 0.661[0.410, 0.999] | 0.065 | 0.725[0.448, 1.106] | 0.2 | |
| P of trend: 0.018* | P of trend: 0.024* | P of trend: 0.093 | |||||
| Direct bilirubin (categorical) | Q1 | – | – | – | – | – | – |
| Q2 | 0.571[0.174, 1.804] | 0.3444 | 0.523[0.155, 1.683] | 0.2832 | 0.588[0.16, 2.074] | 0.4123 | |
| Q3 | 0.444[0.138, 1.359] | 0.1608 | 0.474[0.142, 1.511] | 0.2119 | 0.649[0.179, 2.292] | 0.5024 | |
| Q4 | 0.176[0.035, 0.683] | 0.0189* | 0.18[0.035, 0.702] | 0.0209* | 0.25[0.047, 1.045] | 0.0715 | |
| FBG (continuous) | 3.422[1.116, 11.539] | 0.0370* | 3.099[0.984, 10.647] | 0.0600 | 3.132[0.897, 12.002] | 0.081 | |
| P of trend: 0.078 | P of trend: 0.110 | P of trend: 0.120 | |||||
| FBG (categorical) | Q1 | – | – | – | – | – | – |
| Q2 | 1.105[0.342, 3.574] | 0.8657 | 1.035[0.312, 3.426] | 0.9541 | 1.28[0.347, 4.872] | 0.7103 | |
| Q3 | 0.808[0.185, 3.13] | 0.7625 | 0.779[0.177, 3.046] | 0.7257 | 0.87[0.18, 3.826] | 0.8554 | |
| Q4 | 3.5[1.093, 11.974] | 0.0387* | 3.134[0.943, 11.043] | 0.0665 | 3.482[0.948, 13.86] | 0.0653 | |
| Total protein (continuous) | 1.094[1.012, 1.198] | 0.040* | 1.087[1.008, 1.192] | 0.057 | 1.055[0.984, 1.163] | 0.2 | |
| P of trend: 0.300 | P of trend: 0.400 | P of trend: 0.800 | |||||
| Total protein (categorical) | Q1 | – | – | – | – | – | – |
| Q2 | 1.15[0.324, 3.907] | 0.8237 | 1.008[0.273, 3.526] | 0.9904 | 0.478[0.103, 1.958] | 0.3199 | |
| Q3 | 0.863[0.249, 2.817] | 0.8085 | 0.837[0.237, 2.785] | 0.7745 | 0.446[0.108, 1.67] | 0.243 | |
| Q4 | 2.091[0.675, 6.633] | 0.2021 | 1.911[0.597, 6.22] | 0.2748 | 0.881[0.217, 3.404] | 0.8552 | |
| ALT (continuous) | 0.918[0.826, 1.001] | 0.076 | 0.907[0.814, 0.992] | 0.050 | 0.898[0.796, 0.993] | 0.055 | |
| P of trend: 0.016* | P of trend: 0.010* | P of trend: 0.012* | |||||
| ALT (categorical) | Q1 | – | – | – | – | – | – |
| Q2 | 1.163[0.411, 3.298] | 0.7746 | 1.072[0.361, 3.165] | 0.8989 | 0.976[0.301, 3.121] | 0.9672 | |
| Q3 | 0.339[0.068, 1.314] | 0.1416 | 0.317[0.062, 1.255] | 0.1241 | 0.356[0.065, 1.541] | 0.1912 | |
| Q4 | 0.239[0.049, 0.888] | 0.0467* | 0.194[0.037, 0.768] | 0.03* | 0.138[0.019, 0.67] | 0.0247* | |
Notes: Q1, 0%-25% quantile; Q2, 25%-50% quantile; Q3, 50%-75% quantile; Q4, 75%-100% quantile. Model 1: unadjusted; Model 2: adjusted for age, BMI; Model 3: adjusted for age, BMI, CA125, HE4. * P < 0.05.
Nonlinear Relationship Between Metabolic and Outcome Indicators
Restricted cubic spline models were constructed to analyze the potential nonlinear relationship between metabolic indicators and endometriosis severity. The results showed that, with the exception of FBG which showed a significant nonlinear relationship (P-nonlinear: 0.0362), the remaining metabolic markers did not exhibit a significant nonlinear association with the outcome measures (P-nonlinear > 0.05) (Figures 1 and 2).
Figure 1.
RCS cubic spline plots of FBG (A–C), TP (D–F), and DBIL (G–I) in different models. The vertical dotted line indicates the value of the metabolic indicator when the OR is equal to 1.
Abbreviations: TP, total protein; DBIL, direct bilirubin.
Figure 2.
RCS cubic spline plots of TBil (A–C) and ALT (D–F) in different models. The vertical dotted line indicates the value of the metabolic indicator when the OR is equal to 1.
Predictive Power of Metabolic Indicators
We conducted ROC curve analysis for metabolic indicators that were statistically significant in univariate analyses, including TBil, direct bilirubin, FBG, total protein, and ALT, and computed the AUC values (Figure 3). The AUC ranges from 0 to 1, with 0.5 indicating a random classifier and 1 representing a perfect classifier. The AUC results were as follows: TBil (continuous: 0.645 [0.529, 0.76]; categorical: 0.664 [0.552, 0.776]), direct bilirubin (continuous: 0.644 [0.528, 0.76]; categorical: 0.651 [0.539, 0.762]), FBG (continuous: 0.604 [0.474, 0.733]; categorical: 0.631 [0.512, 0.75]), ALT (continuous: 0.624 [0.506, 0.742]; categorical: 0.652 [0.543, 0.761]). All AUC values were above 0.6, suggesting these indicators possess a high level of predictive capability.
Figure 3.
Results of ROC analysis of metabolic indicators. (A) TBil; (B) Direct bilirubin; (C) FBG; (D) Total protein; (E) ALT.
Discussion
This study explored the relationship between standard metabolic indices and the severity of endometriosis. Findings indicated that ALT exhibited a negative association with endometriosis severity. The RCS analysis showed that the majority of these metabolic indicators bore a substantially nonlinear relationship with outcomes.
The present study showed that CA-125 was positively correlated with the severity of endometriosis. Izabela Kokot et al compared serum inflammatory markers between patients with endometriosis and those without and found that CA-125 concentration was significantly elevated in individuals with endometriosis when compared to the non-endometriosis group (p < 0.001). Another study showed that the AUC for the diagnostic ability of serum IL-32 for endometriosis was 0.638; however, when the serum IL-32 level was combined with the serum CA-125 level, the AUC increased to 0.749, suggesting that CA-125 may improve the accuracy of diagnosing endometriosis.17 It has also been shown that increased CA-125 is a marker of severe and deep infiltrating endometriosis. In addition, many studies reported similar results to our research outcomes.18–21 In addition, HE4 (continuous) was not significantly correlated with endometriosis severity. However, HE4 within the Q2 (25%–50% quantile) range was associated with lower severity compared with Q1 (0%–25% quantile). Previous studies have reported that HE4 alone has limited correlation with endometriosis severity, future studies with larger sample sizes are needed to further explore this relationship.
In terms of hepatobiliary metabolic indicators, our results showed that direct bilirubin, TBil and ALT were significantly associated with endometriosis severity. Bilirubin is a breakdown product of heme, released from the lysis of red blood cells. Slightly elevated plasma bilirubin levels have been associated with protective effects against a range of pathologies, and slight decreases in serum bilirubin concentrations have been linked to an increased risk of cardiovascular and metabolic diseases.22,23 However, few studies have directly elucidated the association between direct bilirubin and endometriosis severity. Shogo Imanaka et al reported higher blood bilirubin levels in endometriotic patients compared to non-endometriotic individuals,24 but the study did not correlate bilirubin levels with the severity of endometriosis. Concurrently, a study by the same agency investigated the effects of iron-related compounds and bilirubin on redox homeostasis in endometriosis and its potential for malignant transformation, revealing higher levels of total iron, heme iron, free iron, and bilirubin in endometriosis patients compared to those with endometriosis-associated ovarian cancer.25 This suggests that low bilirubin may indicate disease progression and malignancy in endometriosis, although further high-quality research is needed to confirm this association. In addition to CA-125 and bilirubin, ALT is also an independent factor. ALT is regarded as a marker of liver injury, and its decreased concentration is generally considered to be of no clinical significance. However, no study has analyzed the relationship between ALT and the severity of endometriosis; therefore, further studies are needed.
Analyzing the correlation between metabolic indices and endometriosis severity could provide clinicians with non-invasive biomarkers for early detection and more accurate monitoring of disease progression. This would enable more timely identification of patients at risk for severe manifestations. Additionally, understanding these metabolic associations may facilitate the development of targeted therapies tailored to specific metabolic profiles, enhancing treatment efficacy and personalization. Ultimately, such insights could improve diagnostic accuracy and guide more effective management strategies, offering significant benefits in the clinical care of endometriosis patients.
The present study explored the relationship between standard metabolic indices and the severity of endometriosis. It was found that ALT was statistically associated with endometriosis severity. The ROC curve analysis showed that these indicators have a robust discriminatory capacity. Nevertheless, this study has some limitations. First, the retrospective design limited the types of data that could be collected, and certain elusive endogenous metabolites were beyond the scope of this study. Second, the inability to capture data across all menstrual cycle phases (proliferative, secretory, and menstrual) is a constraint of this retrospective approach. Third, the sample size of this study is relatively small. Future studies with larger sample sizes are needed to further validate these findings. These limitations are exactly what our next research intends to remedy.
Conclusion
CA-125 and HE4 were identified as significant independent factors affecting the severity of endometriosis. ALT demonstrated a negative correlation with endometriosis severity and emerged as an independent factor with statistical significance. In contrast, FBG, total protein, direct bilirubin and TBil were not found to be independent factors influencing the severity of endometriosis. The logistic regression model incorporating the aforementioned indicators exhibited strong discriminatory power. Future prospective studies with larger samples and more refined designs are needed to further validate these findings.
Funding Statement
No funding was received for this research.
Data Sharing Statement
The data that support the findings of this study are available from either corresponding author, Hongyang Xiao or Ruiqin Tu, upon reasonable request.
Ethics Approval
This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Zhongshan Hospital (Xiamen), Fudan University (No. B2022-046).Written informed consent was obtained from the participants.
Consent to Participate
Written informed consent was obtained from the participants.
Disclosure
The authors report there are no competing interests to declare for this work.
References
- 1.Taylor HS, Kotlyar AM, Flores VA. Endometriosis is a chronic systemic disease: clinical challenges and novel innovations. Lancet. 2021;397(10276):839–852. doi: 10.1016/S0140-6736(21)00389-5 [DOI] [PubMed] [Google Scholar]
- 2.Koninckx PR, Fernandes R, Ussia A, et al. Pathogenesis Based Diagnosis and Treatment of Endometriosis. Front Endocrinol. 2021;12:745548. doi: 10.3389/fendo.2021.745548 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dunselman GA, Vermeulen N, Becker C, et al. ESHRE guideline: management of women with endometriosis. Hum Reprod. 2014;29(3):400–412. doi: 10.1093/humrep/det457 [DOI] [PubMed] [Google Scholar]
- 4.Falcone T, Flyckt R. Clinical Management of Endometriosis. Obstet Gynecol. 2018;131(3):557–571. doi: 10.1097/AOG.0000000000002469 [DOI] [PubMed] [Google Scholar]
- 5.Chapron C, Marcellin L, Borghese B, Santulli P. Rethinking mechanisms, diagnosis and management of endometriosis. Nat Rev Endocrinol. 2019;15(11):666–682. doi: 10.1038/s41574-019-0245-z [DOI] [PubMed] [Google Scholar]
- 6.Ballweg ML. Treating endometriosis in adolescents: does it matter? J Pediatr Adolesc Gynecol. 2011;24(5 Suppl):S2–6. doi: 10.1016/j.jpag.2011.07.003 [DOI] [PubMed] [Google Scholar]
- 7.Kobayashi H, Kawahara N, Ogawa K, Yoshimoto C. A relationship between endometriosis and obstetric complications. Reprod Sci. 2020;27(3):771–778. doi: 10.1007/s43032-019-00118-0 [DOI] [PubMed] [Google Scholar]
- 8.Guidozzi F. Endometriosis-associated cancer. Climacteric. 2021;24(6):587–592. doi: 10.1080/13697137.2021.1948994 [DOI] [PubMed] [Google Scholar]
- 9.Bulun SE, Wan Y, Matei D. Epithelial Mutations in Endometriosis: link to Ovarian Cancer. Endocrinology. 2019;160(3):626–638. doi: 10.1210/en.2018-00794 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hermens M, van Altena AM, Bulten J, van Vliet H, Siebers AG, Bekkers RLM. Increased incidence of ovarian cancer in both endometriosis and adenomyosis. Gynecol Oncol. 2021;162(3):735–740. doi: 10.1016/j.ygyno.2021.07.006 [DOI] [PubMed] [Google Scholar]
- 11.Ac SDS, Capek S, Amrami KK, Spinner RJ. Neural involvement in endometriosis: review of anatomic distribution and mechanisms. Clin Anat. 2015;28(8):1029–1038. doi: 10.1002/ca.22617 [DOI] [PubMed] [Google Scholar]
- 12.Jerman LF, Hey-Cunningham AJ. The role of the lymphatic system in endometriosis: a comprehensive review of the literature. Biol Reprod. 2015;92(3):64. doi: 10.1095/biolreprod.114.124313 [DOI] [PubMed] [Google Scholar]
- 13.Conroy I, Mooney SS, Kavanagh S, et al. Pelvic pain: what are the symptoms and predictors for surgery, endometriosis and endometriosis severity. Aust N Z J Obstet Gynaecol. 2021;61(5):765–772. doi: 10.1111/ajo.13379 [DOI] [PubMed] [Google Scholar]
- 14.Vanni VS, Villanacci R, Salmeri N, et al. Publisher Correction: concomitant autoimmunity may be a predictor of more severe stages of endometriosis. Sci Rep. 2021;11(1):17715. doi: 10.1038/s41598-021-97506-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kim B-S, Kim B, Yoon S, et al. Warburg-like Metabolic Reprogramming in Endometriosis: from Molecular Mechanisms to Therapeutic Approaches. Pharmaceuticals. 2025;18(6):813. doi: 10.3390/ph18060813 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Datkhayeva Z, Iskakova A, Mireeva A, et al. The Multifactorial Pathogenesis of Endometriosis: a Narrative Review Integrating Hormonal, Immune, and Microbiome Aspects. Medicina. 2025;61(5):811. doi: 10.3390/medicina61050811 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Choi YS, Kim S, Ys O, Cho S, Hoon KS. Elevated serum interleukin-32 levels in patients with endometriosis: a cross-sectional study. Am J Reprod Immunol. 2019;82(2):e13149. doi: 10.1111/aji.13149 [DOI] [PubMed] [Google Scholar]
- 18.Fiala L, Bob P, Raboch J. Oncological markers CA-125, CA 19-9 and endometriosis. Medicine. 2018;97(51):e13759. doi: 10.1097/MD.0000000000013759 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Oliveira MAP, Raymundo TS, Soares LC, Pereira TRD, Demoro AVE. How to Use CA-125 More Effectively in the Diagnosis of Deep Endometriosis. Biomed Res Int. 2017;2017:9857196. doi: 10.1155/2017/9857196 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shin KH, Kim HH, Kwon BS, Suh DS, Joo JK, Kim KH. Clinical Usefulness of Cancer Antigen (CA) 125, Human Epididymis 4, and CA72-4 Levels and Risk of Ovarian Malignancy Algorithm Values for Diagnosing Ovarian Tumors in Korean Patients With and Without Endometriosis. Ann Lab Med. 2020;40(1):40–47. doi: 10.3343/alm.2020.40.1.40 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dochez V, Caillon H, Vaucel E, Dimet J, Winer N, Ducarme G. Biomarkers and algorithms for diagnosis of ovarian cancer: CA125, HE4, RMI and ROMA, a review. J Ovarian Res. 2019;12(1):28. doi: 10.1186/s13048-019-0503-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hinds Jr TD, Stec DE. Bilirubin Safeguards Cardiorenal and Metabolic Diseases: a Protective Role in Health. Curr Hypertens Rep. 2019;21(11):87. doi: 10.1007/s11906-019-0994-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hinds Jr TD, Stec DE. Bilirubin, a Cardiometabolic Signaling Molecule. Hypertension. 2018;72(4):788–795. doi: 10.1161/HYPERTENSIONAHA.118.11130 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Imanaka S, Yamada Y, Kawahara N, Kobayashi H. A delicate redox balance between iron and heme oxygenase-1 as an essential biological feature of endometriosis. Arch Med Res. 2021;52(6):641–647. doi: 10.1016/j.arcmed.2021.03.006 [DOI] [PubMed] [Google Scholar]
- 25.Shigetomi H, Imanaka S, Kobayashi H. Effects of iron-related compounds and bilirubin on redox homeostasis in endometriosis and its malignant transformations. Horm Mol Biol Clin Investig. 2021;43(2):187–192. doi: 10.1515/hmbci-2021-0065 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from either corresponding author, Hongyang Xiao or Ruiqin Tu, upon reasonable request.