Association between NHHR and endometriosis and the mediating role of the CALLY index: a cross-sectional study based on NHANES

Table  1 presents the baseline characteristics of participants stratified by NHHR quartiles (Q1 < 1.7606, Q2 1.7606–2.3889, Q3 2.3889–3.1886, Q4 ≥ 3.1886). A total of 3994 individuals were enrolled, representing 63,262,536 individuals after applying sampling weights. The median age of the enrolled individuals was 35 years, and EMS prevalence was 7.19% ( n  = 287 cases out of 3,994). In comparison to the lowest NHHR quartile, individuals in higher quartiles tended to have greater odds of HP and diabetes, be NHW, obese, older, have limited education levels, lower PIR, as well as lower dietary intake of protein, TF, and cholesterol. Table 1 Weighted baseline characteristics of the study population Characteristic N 1 Overall Q1(<1.7606) Q2(1.7606–2.3889) Q3(2.3889–3.1886) Q4(≥3.1886) p -value 3 N  = 63,262,536 2 N  = 16,160,014 2 N  = 15,583,987 2 N  = 15,318,540 2 N  = 16,199,996 2 BMI (Kg/m2) 3,994 27(23,32) 23(21,27) 25(22,30) 28(24,33) 31(26,36) < 0.001 Age (year) 3994 38(29,46) 36(26,45) 37(28,45) 38(29,46) 41(32,47) < 0.001 Race 3,994 0.076  Mexican American 889(8.3%) 175(6.6%) 213(8.1%) 252(9.5%) 249(9.1%)  Non-Hispanic White 1,933(69%) 486(69%) 480(70%) 458(68%) 509(71%)  Non-Hispanic Black 829(12%) 254(14%) 217(13%) 198(12%) 160(9.4%)  Other races 343(10%) 83(10%) 88(9.5%) 91(11%) 81(10%) Education Level 3994 < 0.001  Less Than 9th Grade 309(3.7%) 47(2.2%) 81(4.8%) 92(3.9%) 89(4.0%)  9-11th Grade (Includes 12th grade with no diploma) 591(11%) 130(8.5%) 129(8.6%) 150(12%) 182(14%)  High School Grad/GED or Equivalent 869(22%) 171(17%) 199(19%) 233(25%) 266(29%)  Some College or AA degree 1,324(36%) 325(34%) 350(38%) 323(34%) 326(36%)  College Graduate or above 901(27%) 325(39%) 239(29%) 201(25%) 136(16%) Marital Status 3994 0.004  Married 2,248(58%) 523(54%) 563(58%) 591(60%) 571(58%)  Widowed 58(1.5%) 12(1.1%) 12(1.3%) 13(1.8%) 21(1.9%)  Divorced 349(10%) 68(8.7%) 85(9.9%) 89(9.7%) 107(12%)  Separated 158(3.3%) 40(3.5%) 41(3.1%) 36(3.2%) 41(3.3%)  Never married 802(19%) 264(25%) 201(20%) 187(17%) 150(14%)  Living with partner 379(8.5%) 91(7.8%) 96(7.9%) 83(7.5%) 109(11%) Family Income 3,994 3.01(1.50,4.92) 3.46(1.78,5.00) 3.04(1.53,5.00) 2.94(1.39,4.64) 2.61(1.26,4.47) < 0.001 Alcohol user 3,994 < 0.001  No 1,547(32%) 335(26%) 380(32%) 391(32%) 441(39%)  Yes 2,447(68%) 663(74%) 618(68%) 608(68%) 558(61%) Hypertension 3,994 < 0.001  No 2,994(73%) 812(81%) 767(76%) 739(72%) 676(64%)  Yes 1,000(27%) 186(19%) 231(24%) 260(28%) 323(36%) Diabetes mellitus 3,994 < 0.001  No 3,755(95%) 973(98%) 957(96%) 929(95%) 896(91%)  Yes 239(5.2%) 25(2.1%) 41(4.4%) 70(5.2%) 103(9.0%) C-reactive protein(mg/dL) 3,994 0.23(0.07,0.56) 0.10(0.04,0.31) 0.18(0.06,0.52) 0.28(0.10,0.60) 0.42(0.20,0.76) < 0.001 Creatinine, urine (mg/dL) 3,994 103(56,156) 94(42,144) 105(59,157) 105(57,162) 110(63,160) 0.001 Albumin, urine (ug/mL) 3,994 7(3,13) 6(3,13) 7(3,12) 7(3,13) 7(4,15) 0.002 White blood cell count (1000 cells/uL) 3,994 7.30(5.90,8.80) 6.80(5.50,8.10) 7.00(5.90,8.40) 7.40(6.00,8.80) 7.90(6.60,9.40) < 0.001 Lymphocyte percent (%) 3,994 30(25,35) 30(24,35) 29(25,35) 30(25,34) 30(25,35) 0.376 Monocyte percent (%) 3,994 7.10(5.90,8.30) 7.40(6.20,8.70) 7.20(5.90,8.40) 7.00(5.90,8.10) 6.60(5.60,8.00) < 0.001 Segmented neutrophils percent (%) 3,994 60(54,66) 60(53,66) 60(54,65) 60(55,66) 60(54,65) 0.676 Eosinophils percent (%) 3,994 2.00(1.30,3.10) 2.00(1.20,3.00) 2.10(1.30,3.20) 2.00(1.30,2.90) 2.10(1.40,3.20) 0.092 Basophils percent (%) 3,994 0.60(0.40,0.80) 0.60(0.40,0.90) 0.60(0.40,0.90) 0.60(0.40,0.80) 0.60(0.40,0.80) 0.226 Hemoglobin (g/dL) 3,994 13.70(12.90,14.30) 13.50(12.90,14.10) 13.60(12.90,14.30) 13.80(12.90,14.40) 13.80(13.10,14.50) < 0.001 Albumin (g/dL) 3,994 4.20(4.00,4.40) 4.20(4.00,4.40) 4.20(4.00,4.40) 4.20(3.90,4.40) 4.10(3.90,4.30) < 0.001 Alanine aminotransferase ALT (U/L) 3,994 18(15,23) 17(14,21) 17(14,23) 19(15,24) 20(16,27) < 0.001 Aspartate aminotransferase AST (U/L) 3,994 21(18,24) 21(18,24) 20(18,24) 20(18,24) 21(18,25) 0.07 Blood urea nitrogen (mg/dL) 3,994 10(8,13) 10(8,13) 10(8,13) 10(8,13) 11(8,13) 0.033 Creatinine (mg/dL) 3,994 0.80(0.70,0.80) 0.80(0.70,0.80) 0.80(0.70,0.80) 0.70(0.70,0.80) 0.70(0.70,0.80) 0.613 Bilirubin, total (mg/dL) 3,994 0.60(0.50,0.80) 0.60(0.50,0.80) 0.60(0.50,0.80) 0.60(0.50,0.70) 0.60(0.50,0.70) < 0.001 Alkaline phosphotase (U/L) 3,994 60(49,75) 53(45,64) 58(48,71) 62(52,76) 69(56,84) < 0.001 Protein intake from food (gm) 3,994 68(50,89) 71(52,92) 70(52,91) 67(47,88) 64(48,86) 0.008 Cholesterol intake from food (gm) 3,994 192(117,323) 200(116,333) 200(121,342) 191(120,310) 179(111,307) 0.136 Total fat intake from food (gm) 3,994 69(48,94) 69(49,97) 69(48,96) 70(48,92) 66(47,91) 0.324 1 N not Missing (unweighted) 2 Median (Q1, Q3); n (unweighted) (%) 3 Design-based Kruskal Wallis test; Pearson’s X^2: Rao & Scott adjustment Weighted baseline characteristics of the study population 1 N not Missing (unweighted) 2 Median (Q1, Q3); n (unweighted) (%) 3 Design-based Kruskal Wallis test; Pearson’s X^2: Rao & Scott adjustment When treated as a continuous variable, elevated NHHR was related to greater odds of having EMS. In Model 1, every unit rise in NHHR values correlated with 15.3% greater odds of developing EMS (OR = 1.153, 95% CI: 1.022–1.302, P  = 0.022). In Model 2, the relationship slightly attenuated but was still statistically significant (OR = 1.156, 95% CI: 1.011–1.323, P  = 0.035). In Model 3, every unit rise in NHHR values was related to 14.3% higher odds of having EMS, and the relationship was significant (OR = 1.143, 95% CI: 1.001–1.306, P  = 0.049). The results indicated that, when treated as a continuous variable, higher NHHR was significantly associated with the prevalence of EMS (Table  2 ). Table 2 Weighted multivariate logistic analysis of NHHR and endometriosis Exposure Model Ⅰ Model Ⅱ Model Ⅲ OR (95%CI) P value OR (95%CI) P value OR (95%CI) P value NHHR (continuous) 1.153(1.022, 1.302) 0.022 1.156(1.011, 1.323) 0.035 1.143(1.001, 1.306) 0.049 NHHR non-high-density lipoprotein cholesterol, PIR poverty-income ratio, BMI body mass index, CI confidence interval, OR odds ratio Model I: Unadjusted Model II: Adjusted for race, PIR, marital status, BMI, and age Model III: Adjusted for age, race, marital status, PIR, drinking status, hypertension, diabetes, BMI, dietary protein intake, dietary cholesterol intake, and total dietary fat intake Weighted multivariate logistic analysis of NHHR and endometriosis NHHR non-high-density lipoprotein cholesterol, PIR poverty-income ratio, BMI body mass index, CI confidence interval, OR odds ratio Model I: Unadjusted Model II: Adjusted for race, PIR, marital status, BMI, and age Model III: Adjusted for age, race, marital status, PIR, drinking status, hypertension, diabetes, BMI, dietary protein intake, dietary cholesterol intake, and total dietary fat intake RCS analysis (Fig.  2 ) suggested a linear link between elevated NHHR values and greater EMS prevalence (P for overall = 0.0173, P for non-linear = 0.7643). Fig. 2 Restricted cubic spline (RCS) analysis of the association between NHHR and endometriosis Restricted cubic spline (RCS) analysis of the association between NHHR and endometriosis Subgroup and interaction analyses were carried out to analyze the consistency of the link between NHHR and EMS among subgroups. These subgroups were based on AC (yes or no), race (NHB, MA, NHW, or others), age (< 35 or ≥ 35 years), HP (yes or no), diabetes (yes or no), and BMI (underweight/normal weight or overweight/obese). The association between elevated NHHR and greater odds of having EMS was more pronounced among individuals who aged under 35 (OR = 1.333, 95% CI: 1.042–1.704, P  = 0.024), were NHW (OR = 1.169, 95% CI: 1.018–1.342, P  = 0.028), those with underweight or normal BMI (OR = 1.265, 95% CI: 1.096–1.46, P  = 0.002), and without HP (OR = 1.183, 95% CI: 1.006–1.391, P  = 0.042), with diabetes (OR = 2.042, 95% CI: 1.093–3.818, P  = 0.027), and those who consumed alcohol (OR = 1.169, 95% CI: 1.03–1.326, P  = 0.018). To assess heterogeneity across subgroups, interaction terms were incorporated. No significant differences were found among subgroups (all interaction P values were greater than 0.05). This suggested that the relationship between NHHR and EMS did not vary notably by HP, diabetes, AC, race, BMI, or age. These results suggested a consistent relationship between increased NHHR and greater odds of having EMS among different subgroups, indicating its potential applicability in diverse populations (Fig.  3 ). Fig. 3 Subgroup analysis of the association between NHHR and endometriosis Subgroup analysis of the association between NHHR and endometriosis In addition, a mediation analysis was carried out to assess the potential mediating role of the CALLY index in the relationship between NHHR values and EMS prevalence. The CALLY index partially mediated the association between NHHR and EMS, accounting for 13.95% of the effect (proportion mediated = 0.1394502, 95% CI: 0.1347966–0.1445798). A Sankey diagram illustrates the relationship among NHHR, the CALLY index, and EMS (Fig.  4 ). Another Sankey diagram illustrates the results of the sensitivity analysis of the CALLY index as a mediating variable between NHHR and endometriosis. (Fig.  5 ). Fig. 4 CALLY index as a mediator in the association between NHHR and endometriosis CALLY index as a mediator in the association between NHHR and endometriosis Fig. 5 The sensitivity analysis of the CALLY index as a mediating variable between NHHR and endometriosis The sensitivity analysis of the CALLY index as a mediating variable between NHHR and endometriosis

Data utilized in this article were retrieved from NHANES, a cross-sectional survey guided by the National Center for Health Statistics (NCHS). It collected laboratory, dietary, demographic, and health data from American residents to analyze their nutritional and health status. To ensure the representativeness of the sample, this survey employed a complex probability sampling method. Data were collected from participants through standardized questionnaires, interviews, and tests, and informed consent was provided by them. Moreover, the research protocol obtained approval from the Ethics Review Board of NCHS. This cross-sectional study utilized data from three NHANES survey cycles: 2001–2002, 2003–2004, and 2005–2006. A total of 31,509 participants completed demographic surveys, dietary assessments, physical exams, laboratory tests, and health condition questionnaires. Male individuals ( n  = 15,381) were excluded. Furthermore, individuals without data on EMS ( n  = 11,896) or with insufficient information to calculate NHHR ( n  = 220) were excluded. In addition, individuals without data on key covariates ( n  = 18), encompassing diabetes, educational background, hypertension (HP), marital status, and alcohol consumption (AC), were removed. Ultimately, 3994 individuals were included. Figure  1 illustrates the participant selection process. Fig. 1 Flowchart of the screening process from the National Health and Nutrition Examination Survey Flowchart of the screening process from the National Health and Nutrition Examination Survey The primary outcome variable was EMS, and the reproductive health questionnaire in the NHANES was applied to identify EMS. Female individuals were regarded as having EMS if they responded affirmatively to the question, “Have you been diagnosed with EMS by a physician or other healthcare professional?”. The independent exposure variable was NHHR. According to the serum HDL-C and total cholesterol (TC) values obtained from the NHANES 2001–2002, 2003–2004, 2005–2006, it was calculated by dividing non-HDL-C (obtained by subtracting HDL-C from TC) by HDL-C [ 22 ]. The CRP–albumin–lymphocyte (CALLY) index was calculated as: [(albumin, g/L) × (lymphocyte count per 1,000 cells/µL)]/(CRP, mg/dL) [ 23 ]. To further improve the analysis, the CALLY index was log-transformed to enhance normality and reduce the influence of extreme values on the results. Demographic, health, lifestyle, and dietary-related data were collected via the NHANES household interviews, including cholesterol intake, poverty income ratio (PIR), race, AC, protein intake, body mass index (BMI), HP, diabetes, age, total fat (TF) intake, as well as marital status. BMI was calculated by dividing weight (kg) by height (m) squared. Participants were categorized based on BMI as underweight/normal weight (< 25 kg/m²) or overweight/obese (≥ 25 kg/m²). HP was confirmed according to the administration of antihypertensive medication, the self-reported diagnosis of HP, a measured diastolic blood pressure of 80 mmHg or higher, and a systolic blood pressure of 130 mmHg or higher. The cut-off value of 130/80 mmHg was derived from the updated HP guidelines issued by the 2017 American College of Cardiology and American Heart Association [ 24 ]. Following the American Diabetes Association, diabetes was confirmed based on fasting blood glucose of 126 mg/dL or higher, administration of insulin or oral hypoglycemic drugs, HbA1c of 6.5% or higher, or self-reported diagnosis of diabetes [ 25 ]. Race encompassed non-Hispanic Black (NHB), Mexican American (MA), non-Hispanic White (NHW), as well as others [ 26 ]. Variables with missing data greater than 20% were ruled out. Multiple imputation by chained equations with random forests was applied for continuous variables with missing data ≤ 20%. Multiple imputation was performed using the mice package with five iterations (m = 5) (Multiple Imputation by Chained Equations, MICE with random forest). Statistical analyses were conducted following the NHANES analytic guidelines. Examination weights were calculated as WTMEC2YR divided by 3. The analyses also accounted for the complex multistage clustered survey design, including SDMVSTRA and SDMVPSU. Analyses were performed using the survey package in R [ 27 ]. Participants were categorized based on BMI as underweight/normal weight (< 25 kg/m²) or overweight/obese (≥ 25 kg/m²). The Shapiro-Wilk test was utilized to assess the normality of continuous variables. Given that continuous variables were all non-normally distributed, they were represented by the median and interquartile range (IQR), and between-group comparisons were carried out via the Kruskal-Wallis test. Frequencies and percentages were utilized to represent categorical variables, and between-group comparisons were made through weighted chi-square tests. Multicollinearity of variables included in the models was assessed, and variables with a variance inflation factor ≥ 5 were excluded (Supplementary Table 1). Three weighted logistic regression models were applied to analyze the link between NHHR values and EMS prevalence. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. Model 1 was uncontrolled, and Model 2 controlled for race, PIR, marital status, BMI, and age. Model 3 further controlled for diabetes, AC, HP, dietary intake of cholesterol, protein, and TF. Restricted cubic splines (RCS) with four knots (0.05, 0.35, 0.65, and 0.95) were combined with logistic regression models to explore the potential association between NHHR and EMS, with a specific inflection point of 2.4. Subgroup analyses and interaction tests stratified by AC (yes or no), race (NHB, MA, NHW, or others), HP (yes or no), diabetes (yes or no), age (< 35 or ≥ 35), and BMI (< 25 kg/m² as underweight or normal, ≥ 25 kg/m² as overweight or obese) were applied to examine the reliability of the results. To investigate whether the CALLY index mediated the relationship between NHHR and EMS, a causal mediation analysis based on the counterfactual framework was performed using the mediation package in R [ 28 ]. Two models were constructed for the mediation analysis. The mediation model employed a linear regression to evaluate the association between NHHR (exposure) and the CALLY index (mediator). This model adjusted for multiple covariates, including age, race, marital status, PIR, BMI, AC, HP, diabetes, and dietary intake of protein, cholesterol, and TF. The outcome model used a logistic regression including both NHHR (exposure) and the CALLY index (mediator) to assess their independent effects on EMS (outcome), adjusting for the same set of covariates. The magnitude of the mediation effect was quantified as the proportion mediated, calculated as the ratio of the indirect effect to the total effect. Statistical significance of the mediation effect was assessed via bootstrap resampling with 1000 iterations, and bias-corrected CIs for percentiles were used to evaluate the stability and reliability of the indirect effect. Additionally, sensitivity analyses without covariate adjustment were performed for both the linear and logistic regression models. R software (version 4.4.2) was utilized to analyze data, and a P -value less than 0.05 (two-sided) suggested statistical significance. Generalized linear models and mediation analyses were performed using the survey and mediation packages, respectively.

This study revealed a positive linear association between NHHR and EMS, with the CALLY index acting as a partial mediator of this relationship. Active management of inflammation and blood lipid levels may help reduce the prevalence of EMS. To validate these observations, larger-scale prospective cohort studies are warranted. Future research should focus on investigating whether targeted interventions to reduce NHHR can improve clinical outcomes in individuals with EMS, thereby offering novel approaches for preventing and controlling EMS and its related complications.

EMS, a chronic inflammatory disease affecting women of reproductive age, can cause pelvic pain and infertility. Delays in diagnosis and challenges in treatment impose a substantial burden on both quality of life and socioeconomic outcomes [ 3 ]. This study, involving 3994 individuals, identified an independent association between higher NHHR and elevated EMS prevalence. The link between NHHR values and EMS was linear and positive. This research highlighted the complex role of NHHR in EMS and demonstrated that the CALLY index partially mediated the association between NHHR and EMS. These results offer significant evidence of the potential involvement of lipid metabolism in EMS development. NHHR stands as a new lipid marker that incorporates both non-HDL-C and HDL-C. Conventional lipid markers (including HDL-C or LDL-C) reflect only one aspect of lipid status. In contrast, NHHR simultaneously represents anti-atherogenic and atherogenic lipid components. Given its cost-effectiveness and clinical practicality, NHHR has been increasingly recognized for its predictive value across multiple diseases in previous studies. Liu et al. [ 29 ] reported that NHHR demonstrated superior sensitivity and specificity in assessing CVD risk. Based on a large-scale longitudinal cohort, Sheng et al. [ 30 ] reported that NHHR outperformed traditional lipid indicators in assessing diabetes risk. Furthermore, studies by Qi [ 18 ] and Qing [ 31 ] revealed associations between NHHR and suicidal ideation, as well as depression. Their findings reveal a relationship between lipid metabolism and psychological health. Currently, increasing attention has been paid to the relationship between lipid markers and EMS. EMS refers to an estrogen-dependent disorder with a multifactorial etiology involving genetic, hormonal, and environmental factors. Several metabolomics studies have demonstrated that abnormal metabolism of phosphatidylcholine and sphingolipids in EMS patients indicates potential disturbances in lipid metabolism [ 32 – 34 ]. A case-control study identified an association between dyslipidemia and EMS. Their results demonstrated that women with EMS exhibited elevated serum LDL-C and TC values, along with decreased HDL-C concentrations [ 35 ]. Previous research reported abnormal lipid profiles in EMS patients, indicating that higher TC values may act as a major contributor to the EMS risk [ 36 ]. High-fat diets, particularly those rich in saturated fats, can rapidly impair the function of type 3 innate lymphoid cells in the gut, leading to compromised intestinal barrier integrity, dysbiosis, and exacerbated inflammation [ 37 ]. Fritsche et al. further reported that high-fat diets could promote the translocation of endotoxins into the bloodstream, thereby stimulating innate immune cells and triggering a transient postprandial inflammatory response [ 38 ]. In this study, NHHR, as a novel indicator of lipid metabolism, was significantly related to EMS. This relationship may be explained by several underlying mechanisms, encompassing inflammatory responses, immune regulation, and hormonal alterations. Furthermore, the CALLY index partially mediated the relationship between NHHR and EMS. CRP is a highly conserved pentameric protein primarily synthesized in the liver and is widely recognized as a highly sensitive biomarker of inflammation and tissue damage. Beyond its role as a biomarker, CRP actively modulates both physiological and pathophysiological processes in inflammation and autoimmunity [ 39 ]. Agirbasli et al. [ 40 ] reported a positive relationship between the TC/HDL-C ratio and high-sensitivity CRP (hs-CRP) levels. Patients with elevated hs-CRP showed enhanced proinflammatory responses, increased glycolytic activity, and altered immune profiles [ 41 ]. Wadham et al. [ 42 ] demonstrated that HDL, through its oxidized phospholipid components, completely prevented CRP-induced upregulation of inflammatory adhesion molecules, thereby eliminating the proinflammatory activity of CRP and suppressing the inflammatory response. Studies have also shown that high LDL-C exposure activates CD4 + and CD8 + T cells, impairs their mitochondrial respiration, drives cellular metabolism toward glycolysis, and induces the production of inflammatory cytokines and mitochondrial reactive oxygen species [ 43 ]. Albumin regulates inflammation by inhibiting pro-inflammatory cytokines and activating the nuclear factor-κB pathway [ 44 ]. Wei et al. [ 45 ] indicate that inflammatory mediators, encompassing TGF-β, IL-6, TNF-α, and IL-1β, are markedly increased in the peritoneal fluid of EMS patients. The inflammatory microenvironment can interact with endometriotic cells (such as epithelial and stromal cells), thus critically contributing to the progression and persistence of the disease. NHHR may promote and sustain chronic inflammation by increasing the expression of these mediators [ 46 ], potentially accelerating the proliferation and spread of ectopic endometrial tissue. Prior studies demonstrate that lipid synthesis in macrophages can regulate cellular energy expenditure, phagocytosis, and the production of inflammatory cytokines. Moreover, intermediates in cholesterol biosynthesis are essential regulators of cytokine production and antiviral responses, and lipid metabolic disorders are linked to abnormal macrophage function [ 47 ]. These findings suggest that elevated NHHR may exacerbate local inflammation by modulating macrophage polarization, thereby promoting the progression of ectopic lesions and contributing to the pathophysiology of EMS [ 48 ]. Existing studies reveal that persistent chronic inflammation may impair estrogen regulation and endometrial receptivity in women [ 49 , 50 ]. Estrogen receptors can further modulate relevant signaling pathways and recruit macrophages to accelerate the proliferation and clonal expansion of endometrial cells, thereby facilitating the initiation and progression of EMS [ 51 ]. Finally, NHHR may contribute to EMS by interfering with hormone metabolism. Wahl et al. [ 52 ] reported notable changes in lipid profiles among EMS patients during hormone therapy. This finding suggests a potential role for NHHR in hormone-related mechanisms. The CALLY index, composed of CRP, serum albumin, and lymphocytes, reflects both inflammatory and nutritional status. Inflammation and immunity play crucial roles in the development and progression of EMS. Moreover, many studies have demonstrated the prognostic and predictive value of the CALLY index in various diseases, including colorectal cancer, myocardial infarction, and gastric cancer [ 53 – 55 ]. Therefore, evaluating women with elevated NHHR levels and decreased CALLY index in clinical practice, followed by effective interventions to mitigate this relationship, may serve as an important strategy for the prevention and management of EMS. The race-stratified subgroup analysis revealed a higher EMS prevalence in NHW women. Several factors may contribute to this difference, such as differential access to healthcare, diagnostic bias, and systemic racism. Bougie et al. [ 56 ] demonstrated that White women exhibited notably greater odds of having EMS in comparison to Black and Hispanic women. Notably, this difference might be attributed to the fact that White patients typically have better access to higher-quality healthcare resources, and their symptom reports are more likely to be taken seriously by clinicians. Schulman et al. [ 57 ] demonstrated that both race and gender may unconsciously influence referral patterns and clinical care. Prior research [ 58 ] demonstrates that Hispanic and Black individuals have a lower likelihood of being diagnosed with EMS, undergoing surgical treatment, or receiving adequate pain management. These disparities may confound the accurate assessment of disease prevalence across racial groups. Farland et al. [ 59 ] suggested that delays in diagnosis may be driven by various factors, including inconsistent symptom recognition by patients and the lack of diagnostic techniques that are non-invasive. However, the disparities in diagnosis tend to decrease once individuals access clinical care. Thus, further exploration is needed to analyze the association between race and EMS. Data utilized in the present article were derived from NHANES, a nationally representative database with a substantial sample size. By incorporating appropriate sampling weights, adjusting for multiple potential confounders, and employing logistic regression, RCS models, and subgroup analyses, this study conducted a comprehensive statistical assessment to enhance the reliability of the results. Nevertheless, some limitations must be acknowledged. First, causality between NHHR and EMS cannot be determined because of the cross-sectional design, which relied on data collected at a single point in time. Hence, the causal relationship needs to be clarified by future prospective studies. Secondly, the applicability of our conclusions to broader populations requires further validation, as the sample was limited to the American population. Thirdly, since NHHR was assessed based on a single baseline measurement, the impact of NHHR variations on EMS risk during the follow-up period remains unclear. Fourthly, EMS was identified based on self-reported physician diagnoses from questionnaires, which may introduce recall bias or lead to underestimation of disease prevalence. Fifthly, due to the limitations of the included population, the relatively small sample size of certain subgroups, such as diabetes, may limit the statistical power of subgroup analyses and interaction tests. This may affect the results of regression analysis, increasing the instability of the results. Therefore, the interpretation of the results of these subgroup analyses should be approached with caution, and their conclusions need to be further validated in a larger prospective study. Finally, although extensive adjustments were made for multiple confounders, the analysis did not include detailed information on blood lipid-lowering medications such as statins, as well as gynecologic and reproductive variables, including parity, contraceptive or hormonal therapy, and menstrual cycle characteristics. These factors may represent important potential confounders, and future studies should incorporate more comprehensive covariate control to validate our findings.

Endometriosis (EMS), as a chronic, estrogen-dependent inflammatory illness, is marked by the presence of endometrial-like tissue (including stroma and glands) outside the uterus. The main clinical manifestations of EMS include deep dyspareunia, menstrual irregularities, frequent urination, chronic pelvic pain, infertility, and dysmenorrhea [ 1 – 3 ]. Nearly 10% of females in their reproductive years are diagnosed with EMS worldwide [ 4 ], significantly impairing their reproductive potential and quality of life [ 5 ]. As the definitive diagnosis of EMS requires an invasive hysteroscopic procedure, a considerable diagnostic delay commonly occurs following the onset of symptoms [ 6 ]. Therefore, identifying factors influencing EMS and developing preventive strategies are essential for improving early detection and management of EMS. Emerging evidence suggests that inflammation, oxidative stress, endocrine and metabolic disturbances, and immune dysregulation may contribute to the pathophysiology of EMS [ 7 ]. Current studies demonstrate that EMS might be linked to lipid metabolism disorders [ 8 , 9 ]. Previous studies have shown that dietary fat modifications can significantly influence the plasma lipid profile, thereby affecting the risk of cardiometabolic diseases [ 10 ]. The Mediterranean diet, characterized primarily by monounsaturated fatty acids (olive oil) and n-3 polyunsaturated fatty acids (fish and nuts), can increase high-density lipoprotein cholesterol (HDL-C), decrease low-density lipoprotein cholesterol (LDL-C), and reduce plasma concentrations of inflammatory markers such as C-reactive protein (CRP) and interleukin-6 [ 11 ]. Ghasemisedaghat et al. [ 12 ] found that a higher fertility diet score, characterized by high intake of vegetables, fruits, whole grains, and omega-3 fatty acids, along with low consumption of red meat and trans fats, was inversely associated with the risk of EMS. The non-high-density lipoprotein cholesterol (non-HDL-C) encompasses all apolipoprotein B-containing lipoprotein cholesterol, namely very-low-density lipoprotein, lipoprotein(a), chylomicrons, intermediate-density lipoprotein, as well as LDL-C [ 13 ]. As the densest lipoprotein in plasma, HDL-C exerts various physiological functions, including the prevention of atherosclerosis [ 14 ]. The ratio of non-HDL-C to HDL-C (NHHR), which can reflect lipid metabolic status, is a recently developed composite lipid marker [ 15 ]. Existing evidence demonstrates that NHHR may outperform conventional lipid markers in assessing cardiovascular disease (CVD) risk [ 16 ]. Moreover, this indicator has exhibited promising utility in assessing the risk of multiple illnesses, including breast cancer, metabolic syndrome, depression, and nephrolithiasis [ 17 – 21 ]. Currently, studies analyzing the association between NHHR and EMS are scarce. Exploring this association is therefore holds considerable clinical relevance. According to data derived from the National Health and Nutrition Examination Survey (NHANES), the present research seeks to assess the link between NHHR values and the odds of having EMS, and to examine the potential mediating role of CALLY index in this relationship.

Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3.