Association between high-density lipoprotein-related inflammation index and female infertility in US reproductive-aged women: insights from NHANES 2013-2020.

In this study, data from the NHANES, an extensive, cross-sectional, population-oriented investigation targeted at assessing the health and nutritional status of both adult and pediatric populations in the U.S, were leveraged. The program is administered by the National Center for Health Statistics (NCHS). The NHANES data include information from structured questionnaires, physical assessments, household interviews, and laboratory evaluations. All methodologies employed in the program are subject to approval and oversight by the NCHS. The included participants provided their signatures on informed consent forms. Since the NHANES provides publicly accessible data, ethical approval from our Institutional Review Board was not necessary. In the present study, NHANES datasets collected from 2013 to 2020, including 44,960 participants, were utilized. The analysis was restricted to women aged 18–45 years ( N  = 6,502). The exclusion criteria were as follows: (1) missing data on infertility ( N  = 978); (2) history of hysterectomy ( N  = 154); (3) history of bilateral oophorectomy ( N  = 60); (4) missing data on HDL-C or monocyte count ( N  = 329); and (5) missing data on covariates ( N  = 1,406). After data exclusion, 3,575 reproductive-age women were included in the final analysis, including 3,096 women without infertility and 479 women diagnosed with infertility. The comprehensive participant selection procedure is shown in Fig.  1 . Fig. 1 Flowchart showing the selection process of the study population. y: years; HDL: high-density lipoprotein; M: monocyte count Flowchart showing the selection process of the study population. y: years; HDL: high-density lipoprotein; M: monocyte count The MHR, LHR, PHR, and NHR were derived from the following laboratory parameters: lymphocyte count, monocyte count, neutrophil count, and platelet count, and HDL-C concentration in peripheral blood samples. The following formulas were utilized to calculate these ratios: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$MHR=monocyte\;number\left(10^3/\mu L\right)/HDL-C\left(mg/dL\right);$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$LHR=lymphocyte\;number\left(10^3/\mu L\right)/HDL-C\left(mg/dL\right);$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$PHR=platelet\;number\left(10^3/\mu L\right)/HDL-C\left(mg/dL\right);$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$NHR=neutrophil\;number\left(103/\mu L\right)/HDL-C\left(mg/dL\right).$$\end{document} To normalize the distributions, LHR, MHR, NHR, and PHR were log-transformed. These variables were then divided into three equal tertiles (T1, T2, and T3) on the basis of their distribution in the study population. This data-driven categorization enabled trend analyses using logistic regression to explore linear associations. Blood samples were collected in the morning after an 8-hour fast. The samples were subsequently shipped to a laboratory authenticated by the NCHS. The samples for cell counting were collected in ethylenediaminetetraacetic acid (EDTA) tubes and analyzed on a Beckman Coulter UniCel DxH 800 analyzer. The samples for HDL-C testing were centrifuged, frozen at −30 °C, delivered to the University of Minnesota, and analyzed with a Cobas 6000 Chemistry Analyzer. The laboratory techniques are described elsewhere [ 22 ]. Infertility status was assessed utilizing self-reports derived from the Reproductive Health Questionnaire. Women were considered infertile if they answered affirmatively to at least one of the following queries: (1) “Have you ever tried to get pregnancy for at least one year but failed?” or (2) “Have you ever sought medical advice or consultation regarding difficulties in conceiving?” Conversely, participants who replied negatively to both questions were classified as non-infertile [ 23 ]. The present study included the following three covariate categories: (1) demographic variables, such as age, body mass index (BMI, defined as the ratio of weight (kilograms) and height 2 (meters 2 )], race, marital condition, poverty-to-income ratio (PIR, defined as the ratio of family income and poverty threshold), and level of education; (2) lifestyle variables, such as alcohol consumption, smoking habit, vigorous activity, and moderate activity, (3) self-reported variables, such as history of diabetes mellitus (DM), history of hypertension, pelvic infection,, age of first menarche, regularity of menstruation, history of hormonal drug use, and history of pregnancy. The detailed information is shown in Table  1 . Table 1 Baseline characteristics of included women from NHANES 2013 to 2020 Characteristics Total participants Non-infertility Infertility P value Number 3575 3096 479 Age, y 32.00 ± 0.19 31.59 ± 0.19 34.50 ± 0.50 < 0.0001 BMI, kg/m 2 29.47 ± 0.23 29.08 ± 0.24 31.87 ± 0.64 < 0.001 PIR 2.77 ± 0.05 2.74 ± 0.05 2.96 ± 0.11 0.03 HDL, mg/dL 57.15 ± 0.37 57.49 ± 0.39 55.08 ± 1.07 0.04 M, 10 3 /µL 0.56 ± 0.00 0.55 ± 0.00 0.58 ± 0.01 0.03 N, 10 3 /µL 4.56 ± 0.04 4.54 ± 0.05 4.68 ± 0.10 0.24 L, 10 3 /µL 2.35 ± 0.01 2.32 ± 0.02 2.48 ± 0.05 0.003 P, 10 3 /µL 265.21 ± 1.59 264.34 ± 1.70 270.52 ± 3.61 0.12 Log (MHR) −4.64 ± 0.01 −4.65 ± 0.01 −4.56 ± 0.03 0.002 Log (NHR) −2.56 ± 0.01 −2.57 ± 0.01 −2.50 ± 0.03 0.05 Log (LHR) −3.20 ± 0.01 −3.22 ± 0.01 −3.11 ± 0.03 0.004 Log (PHR) 1.54 ± 0.01 1.53 ± 0.01 1.60 ± 0.03 0.02 Age group, y < 0.001  < 35, % 60.64 62.66 48.31  ≥ 35, % 39.36 37.34 51.69 BMI group, kg/m2 < 0.0001  < 25, % 37.29 38.70 28.66  25–30, % 22.91 23.67 18.23  ≥ 30, % 39.80 37.63 53.10 PIR 0.17  PIR ≤ 1.3, % 30.64 31.36 26.25  1.3 < PIR < 3.5, % 32.24 32.03 33.52  PIR ≥ 3.5, % 37.12 36.61 40.23 Educational level 0.56  Below high school, % 9.08 9.11 8.87  High school, % 19.14 18.79 21.28  Above high school, % 71.78 72.09 69.86 Marital status < 0.0001  Never married, % 31.76 34.48 15.09  Living alone, % 8.86 8.58 10.59  Living with partner, % 59.38 56.94 74.33 Race 0.17  Mexican American, % 11.67 11.81 10.86  Non-Hispanic White, % 57.75 56.96 62.60  Non-Hispanic Black, % 12.64 12.79 11.75  Other Hispanic, % 7.41 7.72 5.52  Other races, % 10.52 10.72 9.27 Drinking 0.09  Never, % 11.25 11.64 8.88  Former, % 3.70 3.35 5.83  Mild, % 27.99 28.29 26.19  Moderate, % 29.28 29.52 27.84  Heavy, % 27.78 27.21 31.26 Smoking 0.02  Never, % 69.18 70.32 62.25  Former, % 12.86 12.23 16.67  Now, % 17.96 17.45 21.08 Vigorous activity 0.39  No, % 62.37 61.99 64.70  Yes, % 37.63 38.01 35.30 Moderate activity 0.40  No, % 46.84 46.43 49.33  Yes, % 53.16 53.57 50.67 History of DM 0.03  No, % 89.11 89.52 86.62  Borderline, % 4.91 4.99 4.43  Yes, % 5.98 5.50 8.95 History of hypertension < 0.001  No, % 86.43 87.78 78.13  Yes, % 13.57 12.22 21.87 Age of first menarche (y) 0.25  Younger than 10, % 4.14 3.86 5.86  10–12, % 46.87 46.76 47.58  13–15, % 42.40 42.68 40.67  16 and older, % 6.59 6.70 5.90 Regular menstrual periods 0.23  No, % 7.39 7.06 9.39  Yes, % 92.61 92.94 90.61 Hormonal drug use 0.06  No, % 96.16 96.64 93.21  Yes, % 3.84 3.36 6.79 Pelvic infection < 0.0001  No, % 95.87 96.60 91.42  Yes, % 4.13 3.40 8.58 Ever been pregnant < 0.0001  No, % 34.22 36.97 17.33  Yes, % 65.78 63.03 82.67 BMI body mass index, PIR poverty income ratio, HDL high-density lipoprotein, M monocyte, N neutrophil, L lymphocyte, P platelet, MHR monocyte-to-high-density lipoprotein cholesterol ratio, NHR neutrophil-to-high-density lipoprotein cholesterol ratio, LHR lymphocyte-to-high-density lipoprotein cholesterol ratio, PHR platelet-to-high-density lipoprotein cholesterol ratio, DM  diabetes mellitus; y: years. Continuous variables were represented as weighted means ± standard errors, and categorical variables were indicated as weighted frequencies. Weighted linear regression was implemented to evaluate continuous variables, and the weighted chi-square test was implemented to evaluate categorical variables. P  < 0.05 was established as statistical significance Baseline characteristics of included women from NHANES 2013 to 2020 BMI body mass index, PIR poverty income ratio, HDL high-density lipoprotein, M monocyte, N neutrophil, L lymphocyte, P platelet, MHR monocyte-to-high-density lipoprotein cholesterol ratio, NHR neutrophil-to-high-density lipoprotein cholesterol ratio, LHR lymphocyte-to-high-density lipoprotein cholesterol ratio, PHR platelet-to-high-density lipoprotein cholesterol ratio, DM  diabetes mellitus; y: years. Continuous variables were represented as weighted means ± standard errors, and categorical variables were indicated as weighted frequencies. Weighted linear regression was implemented to evaluate continuous variables, and the weighted chi-square test was implemented to evaluate categorical variables. P  < 0.05 was established as statistical significance Drinking status was classified into five types according to the following criteria: (1) participants who never drank included those who reported consuming fewer than 12 drinks throughout their life; (2) former drinkers included those who reported consuming 12 or more drinks within a year or throughout their life but did not partake in drinking during the previous year; (3) mild drinkers included those who reported consuming 1 drinks daily or engaging in binge drinking less than 2 occasions; (4) moderate drinkers included those who reported consuming at least 2 drinks daily or engaging in binge drinking between 2 and 5 occasions; and (5) heavy drinkers included those who reported consuming at least 3 drinks daily or engaging in binge drinking on 5 or more occasions [ 4 ]. Smoking status was classified into the following three types: (1) women who had smoked < 100 cigarettes over their lifetime but who were not currently smoking were identified as never smokers; (2) women who reported smoking ≥ 100 cigarettes over their lifetime but who were not currently smoking were identified as former smokers; and (3) women who reported smoking ≥ 100 cigarettes over their lifetime and were currently smoking were identified as current smokers [ 4 ]. Vigorous activity is an activity that leads to a significant increase in breathing or heart rate that is sustained for ≥ 10 min without interruption. In contrast, moderate physical activity is described as an activity that leads to a slight increase in breathing or heart rate and is maintained for at least 10 consecutive minutes. DM history was classified into the following three types: (1) DM was characterized by several criteria, including a self-reported diagnosis of DM, a fasting plasma glucose concentration ≥ 7.0 mmol/L, a 2-hour post-oral glucose or random plasma glucose concentration ≥ 11.1 mmol/L, an HbA1c concentration ≥ 6.5%, or administration of oral hypoglycemic medications; (2) borderline DM was identified as having abnormal results on these glucose-related assessments but not meeting the diagnostic criteria for DM; (3) no DM was defined as individuals not classified under the previous two categories [ 24 ]. Hypertension was defined as a systolic blood pressure ≥ 140 mmHg, a diastolic blood pressure ≥ 90 mmHg, administration of hypotensive medications, or a previous diagnosis of the condition [ 24 ]. Pelvic infection diseases were defined as those who received treatment for a pelvic infection or pelvic inflammatory disease [ 25 ]. The regularity of menstruation was characterized by regular periods during the past year [ 26 ]. A history of hormonal drug use was defined as the use of estrogen or progesterone, excluding contraceptive methods or hormonal drug use for infertility treatment. All laboratory and questionnaire data were collected during the same NHANES examination cycle visit, ensuring temporal consistency. Given the intricate sampling methodology employed by the NHANES program, all the statistical evaluations were performed using appropriate sampling weights. The included women were categorized into infertile and non-infertile groups. For descriptive analysis, the continuous variables are presented as weighted means ± standard errors, and the categorical variables are presented as weighted frequencies. To evaluate the differences between two groups, weighted linear regression was used to evaluate continuous variables, whereas the weighted chi-square test was used to evaluate categorical variables. To investigate the relationship between the HDL-related inflammation index and female infertility, weighted multivariable logistic regression analyses were performed using the following three models: (1) Model 1, which was adjusted for no covariate; (2) Model 2, which was adjusted for age, race, PIR, level of education, BMI, and marital status; and (3) Model 3, which incorporated the adjustments made in Model 2 and added adjustments for alcohol consumption, smoking habits, vigorous physical activity, moderate physical activity, personal history of DM, personal history of hypertension, pelvic inflammatory diseases, age at onset of menarche, menstrual regularity, previous use of hormonal drugs, and history of pregnancy. The results are expressed as odds ratios (ORs) accompanied by the corresponding 95% confidence intervals (95% CIs). The Log(MHR) was categorized into tertiles for categorical analysis, with T1 designated as the reference group. Multivariable logistic regression analyses were performed under the three aforementioned models to further evaluate the relationship between Log(MHR) and infertility. Furthermore, smooth curve fitting was applied to explore potential linear relationships between Log(MHR) and the probability of infertility. Penalized spline functions were applied within the logistic regression framework, with infertility (binary outcome: 0 = no or 1 = yes) as the dependent variable and Log(MHR) as the continuous predictor. The model used 3 degrees of freedom, with knot placement determined automatically on the basis of the Akaike information criterion (AIC). This approach allowed visualization of the predicted probability of infertility across Log(MHR) values and assess potential linear or nonlinear associations. Subgroup analyses were performed according to age, BMI, PIR, smoking habit, pelvic infection diseases, history of DM, and pregnancy history. Subgroup analyses were performed using Model 3, with all covariates adjusted, except for the stratification variable. Interaction effects within subgroups were evaluated by incorporating multiplicative interaction terms in the logistic regression models, and the results are presented as the P values for interactions. The stratification criteria were as follows: age was stratified into two ranges (younger than 35 years and 35 years or older); the PIR was stratified into three ranges (≤ 1.3, 1.3 to 3.5, and > 3.5); and BMI was stratified into three ranges (less than 25 kg/m², 25–30 kg/m², and 30 kg/m² or higher). Multiple subgroup analyses were conducted, but no adjustment for multiple comparisons was applied, which increased the risk of Type I error and should be interpreted with caution. R (version 4.2.0) and EmpowerStats ( http://www.empowerstats.com ) were used for statistical analyses. A two-tailed test was employed, and P  < 0.05 was considered to indicate statistical significance.

The present study included a total of 3,575 participants, among whom 479 (13.4%) were diagnosed with infertility. Table  1 presents the detailed characteristics of the involved participants. Compared with non-infertile participants, infertile women were older (34.50 ± 0.50 years old versus 31.59 ± 0.19 years, P  < 0.0001) and more likely to be smokers (37.75% versus 29.68%, P  = 0.02). Infertile women had a higher incidence of pelvic inflammatory diseases than did non-infertile women (8.58% versus 3.40%, P  < 0.0001). Moreover, infertile women were significantly more likely to have experienced previous pregnancies compared with non-infertile women (82.67% versus 63.03%, P  < 0.0001). Notably, compared with non-infertile women, infertile women had significantly lower HDL-C levels (55.08 ± 1.07 mg/dL versus 57.49 ± 0.39 mg/dL, P  = 0.04) but higher Log(MHR) values (−4.56 ± 0.03 versus − 4.65 ± 0.01, P  = 0.002), Log(LHR) values (−3.11 ± 0.03 versus − 3.22 ± 0.01, P  = 0.004), and Log(PHR) values (1.60 ± 0.03 versus 1.53 ± 0.01, P  = 0.02). Table  2 shows the associations between HDL-related inflammatory indices (LHR, PHR, NHR, MHR) and infertility risk in women. In Model 1, no significant association between Log(NHR) and infertility was found; Log(MHR), Log(PHR), and Log(LHR) were positively associated with female infertility (Log(MHR): OR = 1.66, 95% CI: 1.22, 2.24; Log(PHR): OR = 1.58, 95% CI: 1.06, 2.34; Log(LHR): OR = 1.72, 95% CI: 1.18, 2.51). After partial adjustment (Model 2), Log(MHR) and Log(LHR) remained significantly associated with infertility (Log(MHR): OR = 1.59, 95% CI: 1.12, 2.25; Log(LHR): OR = 1.66, 95% CI: 1.09, 2.53). In Model 3, only Log(MHR) was associated with infertility (OR = 1.48, 95% CI: 1.02, 2.14). To explore this association in depth, Log(MHR) was categorized into tertiles (T1, T2, and T3). Compared with the T1 group, the risk of infertility was increased in the T3 group across all the models (OR = 1.69, 95% CI: 1.21, 2.35, P  = 0.002 in Model 1; OR = 1.56, 95% CI: 1.06, 2.31, P  = 0.03 in Model 2; OR = 1.49, 95% CI: 1.03, 2.15, P  = 0.04 in Model 3). Similarly, compared with the T1 group, the T2 group exhibited a significantly increased risk of infertility ( P  < 0.05). Smooth curve fitting analysis supported a linear relationship between increasing Log(MHR) levels and elevated infertility risk (Fig.  2 ). Table 2 Weighted multivariable logistic regression for association between HDL-related inflammation index and female infertility Model 1 Model 2 Model 3 OR (95%CI) P value OR (95%CI) P value OR (95%CI) P value Log(NHR) 1.31(1.00, 1.72) 0.05 1.10(0.84, 1.44) 0.49 1.00(0.74, 1.34) 0.99 Log(LHR) 1.72(1.18, 2.51) 0.01 1.66(1.09, 2.53) 0.02 1.50(0.99, 2.27) 0.05 Log(PHR) 1.58(1.06, 2.34) 0.02 1.27(0.83, 1.95) 0.27 1.14(0.75, 1.73) 0.54 Log(MHR) 1.66(1.22, 2.24) 0.001 1.59(1.12, 2.25) 0.01 1.48(1.02, 2.14) 0.04 Log(MHR) Tertiles  T1 Ref Ref Ref Ref Ref Ref  T2 1.60(1.18, 2.16) 0.003 1.52(1.08, 2.14) 0.02 1.54(1.11, 2.13) 0.01  T3 1.69(1.21, 2.35) 0.002 1.56(1.06, 2.31) 0.03 1.49(1.03, 2.15) 0.04 P for trend 0.003 0.034 0.042 OR odd ratio, CI confidence interval, Ref reference, NHR neutrophil-to-high-density lipoprotein cholesterol ratio, LHR lymphocyte-to- high-density lipoprotein cholesterol ratio, PHR platelet-to-high-density lipoprotein cholesterol ratio, MHR monocyte-to-high-density lipoprotein cholesterol ratio Model 1 unadjusted Model 2 adjusted for age, race, marital status, educational level, PIR, and BMI Model 3 adjusted for all covariates Weighted multivariable logistic regression for association between HDL-related inflammation index and female infertility OR odd ratio, CI confidence interval, Ref reference, NHR neutrophil-to-high-density lipoprotein cholesterol ratio, LHR lymphocyte-to- high-density lipoprotein cholesterol ratio, PHR platelet-to-high-density lipoprotein cholesterol ratio, MHR monocyte-to-high-density lipoprotein cholesterol ratio Model 1 unadjusted Model 2 adjusted for age, race, marital status, educational level, PIR, and BMI Model 3 adjusted for all covariates Fig. 2 Smooth curve fitting showing the association between Log(MHR) and female infertility. Penalized spline functions within a logistic regression model were applied, with infertility as the binary outcome (0 = no, 1 = yes) and Log(MHR) as the continuous predictor. MHR: monocyte-to-high-density lipoprotein cholesterol ratio Smooth curve fitting showing the association between Log(MHR) and female infertility. Penalized spline functions within a logistic regression model were applied, with infertility as the binary outcome (0 = no, 1 = yes) and Log(MHR) as the continuous predictor. MHR: monocyte-to-high-density lipoprotein cholesterol ratio To evaluate the association between Log(MHR) and the risk of infertility in various populations, subgroup analyses was conducted using Model 3, which adjusted for all covariates except the stratification factor. Log(MHR) was analyzed by a continuous variable in Fig.  3 , and the association between Log(MHR) and the risk of infertility was pronounced among females younger than 35 years (OR = 1.79, 95% CI: 1.05, 3.07), those with a PIR between 1.3 and 3.5 (OR = 2.23, 95% CI: 1.16, 4.26), DM-negative females (OR = 1.56, 95% CI: 1.05, 2.32), pelvic inflammatory disease-negative females (OR = 1.54, 95% CI: 1.04, 2.28), and women who had never been pregnant (OR = 3.33, 95% CI: 1.54, 7.26). Furthermore, significant interactions ( P for interactions < 0.05) were observed between Log(MHR) and both age and pregnancy history. Fig. 3 Subgroup analysis of the association between Log(MHR) as a continuous variable and female infertility. Logistic regression (Model 3, adjusted for all covariates except the stratification factor) was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) across different subgroups. y: years; PIR: poverty-to-income ratio; BMI: body mass index; DM: diabetes mellitus Subgroup analysis of the association between Log(MHR) as a continuous variable and female infertility. Logistic regression (Model 3, adjusted for all covariates except the stratification factor) was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) across different subgroups. y: years; PIR: poverty-to-income ratio; BMI: body mass index; DM: diabetes mellitus As shown in Table  3 , Log(MHR) was analyzed by tertiles, and compared with that the T1 group, infertility risk was elevated in the T2 and T3 groups across most subgroups. This trend was evident among females with a PIR of 1.3–3.5 (T2 versus T1: OR = 1.79, 95% CI: 1.03, 3.10; T3 versus T1: OR = 1.90, 95% CI: 1.09, 3.31), DM-negative females (T2 versus T1: OR = 1.45, 95% CI: 1.02, 2.06; T3 versus T1: OR = 1.50, 95% CI: 1.01, 2.24), and females who had never been pregnant (T2 versus T1: OR = 3.05, 95% CI: 1.24, 7.55; T3 versus T1: OR = 3.29, 95% CI: 1.30, 8.31). Table 3 Subgroup analysis of the association between log(MHR) tertiles and female infertility in model 3 Subgroup T1 T2 OR (95%CI) T3 OR (95%CI) P for trend P for interaction Age, y 0.14  = 35 Ref 1.62(0.99, 2.65) 1.35(0.76, 2.38) 0.28 PIR 0.8  PIR < = 1.3 Ref 1.49(0.80, 2.76) 1.55(0.94, 2.56) 0.10  1.3 < PIR  = 3.5 Ref 1.71(0.93, 3.13) 1.40(0.60, 3.31) 0.34 BMI group, kg/m 2 0.08  < 25 Ref 1.55(0.81, 2.94) 1.51(0.80, 2.87) 0.15  25 < = BMI = 30 Ref 0.96(0.56, 1.67) 0.82(0.46, 1.47) 0.43 Smoking 0.50  Never Ref 1.81(1.24, 2.65) 1.57(1.05, 2.35) 0.03  Former Ref 1.69(0.60, 4.78) 1.12(0.46, 2.71) 0.8  Now Ref 1.01(0.47, 2.16) 1.48(0.80, 2.74) 0.13 History of DM 0.68  No Ref 1.45(1.02, 2.06) 1.50(1.01, 2.24) 0.05  Borderline Ref 1.01(0.07, 3.08) 1.18(0.01, 4.29) 0.93  Yes Ref 4.44(1.11, 17.82) 3.25(0.79, 13.33) 0.30 Pelvic infection 0.95  No Ref 1.59(1.17, 2.18) 1.53(1.02, 2.28) 0.05  Yes Ref 1.48(0.12, 4.17) 2.24(0.26, 5.50) 0.29 Ever been pregnant 0.003  No Ref 3.05(1.24, 7.55) 3.29(1.30, 8.31) 0.03  Yes Ref 1.46(1.00, 2.14) 1.36(0.93, 1.98) 0.12 OR odd ratio, CI confidence interval, Ref reference, BMI body mass index, PIR poverty income ratio, DM diabetes mellitus, y years Model 3: age, BMI, PIR, educational level, marital status, race, alcohol intake, smoking, vigorous activity, moderate activity, history of DM, history of hypertension, age of first menarche, regular menstrual periods, hormonal drug use, pelvic infection and ever pregnant were fully adjusted Subgroup analysis of the association between log(MHR) tertiles and female infertility in model 3 OR odd ratio, CI confidence interval, Ref reference, BMI body mass index, PIR poverty income ratio, DM diabetes mellitus, y years Model 3: age, BMI, PIR, educational level, marital status, race, alcohol intake, smoking, vigorous activity, moderate activity, history of DM, history of hypertension, age of first menarche, regular menstrual periods, hormonal drug use, pelvic infection and ever pregnant were fully adjusted

The present study revealed a notable positive association between the MHR and infertility risk in reproductive-age women in the U.S., suggesting that the MHR may reflect underlying inflammatory or metabolic alterations linked to reproductive dysfunction. However, given the observational design, modest effect size, and potential for residual confounding factors, the clinical applicability of the MHR should be examined through future longitudinal studies and mechanistic investigations.

The present investigation revealed a positive association between the MHR and female infertility after adjustment for various confounding variables. However, associations between infertility and the LHR and PHR were attenuated after multivariable adjustment, and the NHR was not significantly related with infertility in any model. A consistent positive linear relationship between the MHR and infertility risk was observed, which was further supported by subgroup analyses. Moreover, age and previous pregnancy history contributed to the heterogeneity of this association. To our knowledge, this is the first large-scale population-based study using the MHR and NHANES data to report a positive association between MHR and infertility in women. These findings suggest that elevated MHR levels may reflect a proinflammatory state linked to reproductive dysfunction. However, because this association is observational, no causal inference should be drawn at this stage. The MHR is an emerging inflammatory biomarker that combines monocyte count and HDL levels, and it was originally used to assess the prognosis of cardiovascular disease [ 27 , 28 ] and the morbidity of chronic kidney disease [ 29 , 30 ]. Although the link between HDL-C and female fertility has been examined in prior studies, the results are inconsistent [ 31 – 34 ]. The present study found the MHR, composed of HDL-C and monocytes, was positively associated with infertility. Monocytes are key sources of proinflammatory cytokines, such as TNF-α, IL-6, and IL-1β [ 35 ], which can disrupt folliculogenesis [ 36 – 38 ], steroidogenesis [ 39 ], and ovulatory signaling pathways [ 40 , 41 ]. Elevated monocyte activation may also impair endometrial receptivity and embryo implantation through local inflammation and oxidative stress [ 42 , 43 ]. Moreover, HDL-C dysfunction has been linked to reduced antioxidant capacity and impaired reverse cholesterol transport, compromising the availability of cholesterol for steroid hormone biosynthesis [ 5 , 6 ]. Dysfunctional HDL may also fail to suppress inflammatory responses in reproductive tissues [ 7 – 9 ]. Therefore, an increased MHR may reflect both heightened inflammation and reduced HDL-mediated protection, hindering reproductive success. Specifically, associations between female infertility and other HDL-related inflammatory indices may be attenuated due to confounding factors, such as age, BMI, marital status, and other reproductive health conditions. These covariates may have a stronger influence on infertility risk than the inflammatory indices alone, thereby diminishing the independent association. In contrast, only the MHR remained significantly associated with infertility, while the association between infertility and the other HDL-related indices was not significant. These findings suggest that fluctuation of monocyte may directly associate role in reproductive dysfunction. However, the degree of association was modest, and confounding from unmeasured variables cannot be excluded. According to subgroup analyses, the association between the MHR and infertility was not significant in women aged ≥ 35 years or in those with previous pregnancies. Age-related decreases in ovarian reserve may overshadow inflammatory effects [ 44 , 45 ], and prior pregnancies (including abortion or cesarean delivery) may independently affect fertility [ 46 , 47 ]. Similar findings have been reported according to analyses of the associations between female infertility and other factors, such as the cardiometabolic index [ 4 ], visceral adipose tissue area [ 48 ], and body fat distribution [ 49 ]. In the subgroups stratified by PIR, only those with a PIR between 1.3 and 3.5 exhibited a significant association between Log(MHR) and infertility, possibly reflecting varying susceptibility to inflammation and healthcare access [ 50 ]. In participants with diabetes or borderline diabetes, the lack of association may be due to the overriding impact of metabolic disease and related treatments (e.g., metformin) on fertility [ 51 , 52 ]. In addition, no significant association between Log(MHR) and infertility was detected in women who had experienced pelvic inflammatory diseases. Pelvic inflammatory disease is considered an independent risk factor for infertility; compared with those without a history, women with a history of pelvic inflammatory disease have a 1.57 times higher incidence of infertility [ 53 ]. These findings highlight the complexity of infertility pathophysiology and underscore the need for careful interpretation. The observed associations are hypothesis-generating associations and must be confirmed in longitudinal and mechanistic studies. The present study had several limitations. First, owing to its cross-sectional design, the temporality between exposure and outcome could not be established, preventing causality analyses; the observed associations may also reflect reverse causation. Second, infertility is a heterogeneous condition encompassing diverse etiologies, such as polycystic ovary syndrome, tubal obstruction, and endometriosis, each with distinct pathophysiological mechanisms. In the present study, infertility was treated as a single binary outcome, which may mask differences in the relationship between the MHR and specific subtypes. In particular, chronic inflammation and lipid metabolism dysregulation may be more relevant in ovulatory disorders than in structural disorders. Third, the NHANES partly relies on self-reported data, and infertility status was not validated through clinical diagnosis or medical records. This may introduce recall or misclassification bias. These limitations have been acknowledged in the literature and are considered inherent to the dataset. Finally, although a significant association between the MHR and female infertility was identified, no clinical cutoff was established for the MHR, and its sensitivity and specificity as a diagnostic tool remain unknown. The predictive value of the MHR remains exploratory, and further studies are needed to establish optimal cutoff points, assess sensitivity and specificity, and determine its applicability in clinical screening and risk stratification.

Female infertility is identified by the failure to conceive after 12 months of routine non-contraceptive coitus in females younger than 35 years, and after 6 months in females aged ≥ 35 years [ 1 ]. Female infertility affects approximately 12.6–17.5% of reproductive-age women worldwide [ 2 ]. As the third most pressing global health issue after cancer and heart disease [ 3 ], infertility imposes socioeconomic burdens and psychological distress [ 4 ]. Identifying modifiable risk factors for infertility is crucial for improving reproductive outcomes. High-density lipoprotein (HDL) plays essential roles in female fertility by mediating lipid transport and facilitates cholesterol delivery for steroid hormone biosynthesis [ 5 ]. Clinically, HDL deficiency is associated with reduced ovarian reserve and impaired oocyte maturation [ 6 ]. In addition, HDL directly interacts with hematopoietic cells, including monocytes, neutrophils, lymphocytes, and platelets, to influence their activation, cytokine secretion, migration, and survival; these interactions enable HDL to shape both innate and adaptive immune responses, contributing to its systemic anti-inflammatory effects [ 7 – 9 ]. HDL attenuates monocyte adhesion, suppresses neutrophil extracellular trap formation, and regulates lymphocyte proliferation by influencing cytokine secretion, cell activation, and trafficking, thereby shaping systemic inflammation [ 10 – 12 ]. Emerging evidence highlights the role of chronic inflammation in female reproductive health. Inflammation is associated with several critical reproductive procedures, including ovulation, embryo implantation, placentation, and labor [ 13 – 16 ]. Dysfunctional inflammation is common in several conditions, such as polycystic ovarian syndrome (PCOS) [ 17 ], endometriosis [ 18 , 19 ], and premature ovarian insufficiency [ 20 ]. Notably, low-dose aspirin can increase conception rates among women experiencing chronic inflammation because of its anti-inflammatory profile [ 21 ]. Given the integrative role of HDL in lipid transport and immune modulation, composite indices such as monocyte-to-HDL-cholesterol (HDL-C, MHR), neutrophil-to-HDL-C (NHR), lymphocyte-to-HDL-C (LHR), and platelet-to-HDL-C (PHR) may better capture inflammation-related reproductive dysfunction than HDL-C alone. These ratios reflect the dynamic balance between proinflammatory cellular activity and anti-inflammatory lipid carriers, potentially offering superior discriminatory power for infertility risk stratification. To explore this hypothesis, data from the National Health and Nutrition Examination Survey (NHANES) were used to assess whether HDL-C–related inflammation indices, such as the MHR, are associated with infertility risk. The reliability of the present results was evaluated by subgroup analyses. Identifying a simple and cost-effective marker for predicting infertility risk could improve our understanding of infertility risk.

Supplementary Material 1. Supplementary Material 1.