Association of metal exposure in follicular fluid with assisted reproductive technology outcomes in East Chinese women.

This study enrolled female patients who underwent ART at the Department of Reproductive Medicine, Affiliated Hospital of Shandong Second Medical University between January and June 2024. Participants were women aged 18 to 40 years. Those excluded from the study had: (1) serious health issues impacting metal metabolism, such as liver or kidney disorders and diabetes; (2) a clear history of exposure to heavy metals or harmful chemicals; (3) recent medication use that could alter metal levels; (4) hydrosalpinx or uterine conditions that hindered embryo implantation; (5) a background of repeated implantation failures, miscarriages, or chromosomal issues. Ultimately, 139 women met the criteria for inclusion. Data collection was performed via the medical records system, encompassing: (1) demographic details; (2) hormone levels, including anti-Müllerian hormone (AMH), follicle stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E 2 ), and progesterone (P); (3) ART outcomes, such as the number of retrieved oocytes, metaphase II (MII) oocytes, two pronuclei (2PN) fertilizations, high-quality embryos, and clinical pregnancy results. The oocyte maturation rate was calculated as the percentage of MII oocytes compared to the total retrieved. High-quality embryos were defined as grade 1–2 embryos derived from 2PN zygotes that developed into either 7–10 cell embryos on day 3 or compacted morulae, according to the Vienna Consensus [ 10 ]. The high-quality embryo rate was calculated as the percentage of high-quality embryos among all cleaved embryos derived from 2PN zygotes. The hospital’s ethics committee granted approval for the study (Approval ID: wyfy-2022-ky-173). Participants undergoing IVF/ICSI treatment received ovarian stimulation with gonadotropins according to standard clinical protocols. Follicular growth and endometrial changes were monitored via serial serum estradiol measurements and transvaginal ultrasonography. When the dominant follicles reached a size of 17 mm or more, human chorionic gonadotropin (hCG) was administered, and oocyte retrieval was performed via transvaginal aspiration within 36 h post-hCG injection. When a single largest follicle was identifiable, it was selected for FF collection. In cases where multiple follicles of similar size were present, the follicle that was aspirated first was chosen for FF collection. After obtaining oocytes from the FF of one follicle, 3–10 mL of FF was collected from the same follicle. FF was subjected to visual inspection. Blood-contaminated samples were discarded, and only clear or slightly straw-colored fluids were retained for subsequent analysis. The qualified samples collected were centrifuged at 2000 rpm for 10 min, and the supernatant was separated, placed into 2 mL cryotubes, and stored at -80 °C for future analysis. Based on clinical requirements, the oocytes were fertilized using IVF/ICSI methods, followed by observation of embryo development. A trace element analyzer (TC-3010 C, Tiancheng Electronic Technology, Jining China) was employed to quantitatively assess eight metal elements. Zn, Fe, Mn, Ca, and Mg were analyzed by polarography; Cd, Pb, and Cu were determined by differential potentiometric stripping analysis. External calibration used gradient-diluted standards, and all calibration curves showed excellent linearity ( R  ≥ 0.999). Next, 20 µL of the FF sample was placed in an element activation tube, left to sit for 5 min, and then centrifuged at 4000 rpm for 1 min to collect the supernatant for analysis. For polarographic analysis, 1 mL each of reagents R1 and R2 was mixed in a 1:1 ratio in the sample cup, followed by the addition of 100 µL of the prepared sample. For metal elements analyzed via the leaching method, the electrode was coated with 1 mL of coating liquid, after which 1 mL of reagent and 20 µL of the prepared sample were added to the sample tube, which was allowed to sit for over 5 min before measurement. Each sample was analyzed in duplicate, and the mean value was used for statistical analysis. Method validation established detection limits of ≤ 1 × 10⁻⁷ mol/L (polarography) and ≤ 0.1 µg/L (stripping analysis). Since Cd was undetectable in all samples, only the remaining seven metals with 100% detection rates were included in statistical analysis, and values below limit of detection (LOD) were imputed as LOD/√2. Method precision showed coefficient of variation (CV) ≤ 1% (polarography) and ≤ 3% (stripping analysis), and accuracy was confirmed through spike-recovery tests (86.9-103.12%). Quality control included daily instrument rinsing, blank correction every 20 samples, and verification using certified control materials with specified target concentrations (Zn: 152.9 µmol/L; Fe: 179.1 µmol/L; Mn: 0.87 µmol/L; Ca: 1.559 mmol/L; Mg: 1.286 mmol/L; Cd: 0.05 µg/L; Pb: 100.00 µg/L; Cu: 19.67 µmol/L). To ensure data integrity, subsequent analyses included only metal elements with a 100% detection rate. Chi-square and Mann-Whitney U tests were utilized to examine demographic variations related to different ART parameters. Given the non-normal distribution of hormone levels, metal concentrations, and ART metrics, the data were expressed as median values along with interquartile ranges (P25–P75). To assess the relationships among metal elements and between these elements and serum hormone levels, Spearman correlation analysis was conducted. The Mann-Whitney U test was also applied to evaluate the differences in metal concentrations across various ART parameter categories. Furthermore, age and body mass index (BMI), AMH, E 2 , and stimulation protocol emerged as significant confounding variables influencing oocyte and embryo quality, which were accounted for using a logistic regression model. To explore the independent relationships between each metal element and ART indicators, metal concentrations were divided into low, medium, and high categories based on tertiles, with the low concentration group acting as the reference. Metal concentrations were categorized into tertiles to facilitate a flexible assessment of potential nonlinearity and to minimize sensitivity to extreme values, while ensuring an adequate sample size within each exposure group, given the size of the cohort. Logistic regression was performed to determine odds ratios (OR) and 95% confidence intervals (CI), while controlling for the same confounding variables. To account for multiple comparisons across all tested associations (42 tests comprising 7 metals, 3 ART outcomes, and 2 exposure comparisons), we applied the Benjamini-Hochberg procedure to control the false discovery rate (FDR). Associations with FDR-adjusted Q values < 0.05 were considered statistically significant. A sensitivity analysis was performed to examine the robustness of the primary findings. The associations between metal exposures and ART outcomes were re-evaluated using logistic regression models with natural log (LN)-transformed continuous metal concentrations, while maintaining adjustment for the same covariates. The RCS model was utilized to effectively model and visualize the potential nonlinear dose-response relationships between continuous metal concentrations and ART parameters. Four knots were strategically placed at the 5th, 35th, 65th, and 95th percentiles of the metal concentration distributions. Lastly, a BKMR model was employed to assess the joint effects, overall mixture impact, and potential interactions among the seven metal elements on ART outcomes. BKMR analyses were performed using the R bkmr package with Gaussian kernel, default priors, and 20,000 MCMC iterations. Model convergence was confirmed by visual inspection of trace plots and Gelman-Rubin diagnostics (all R-hat < 1.1). Statistical analyses were carried out using IBM SPSS Statistics v26.0 and R v4.5.0, with a significance threshold set at two-sided P -values < 0.05.

The age of participants varied from 22 to 39 years, with an average age of 32.05 years and a mean BMI of 24.03 kg/m² (Table  1 ). The rates of oocyte maturation (87.5%) and high-quality embryos (66.66%) were divided into high and low categories based on median values, while clinical pregnancy outcomes were analyzed between those who achieved pregnancy and those who did not. Women in the high-quality embryo category exhibited significantly elevated LH levels compared to those in the low-quality group. The average age of women who did not achieve clinical pregnancy was notably higher than that of those who did. Demographic and other laboratory metrics did not differ significantly between high and low oocyte maturation groups. Table 1 Demographic characteristics of study participants by ART outcomes Characteristics Total participants Oocyte maturation rate (%) P High-quality embryo rate (%) P Clinical pregnancy P ( n  = 139) <87.5 ( n  = 68) ≥ 87.5 ( n  = 71) <66.66 ( n  = 65) ≥ 66.66 ( n  = 74) No ( n  = 68) Yes ( n  = 71) Age (years) 32.05 ± 4.03 32.49 ± 4.10 31.83 ± 3.98 0.341 31.63 ± 4.12 32.61 ± 3.93 0.155 33.03 ± 4.29 31.31 ± 3.61 0.011 BMI (kg/m2) 24.03 ± 3.96 23.91 ± 3.88 24.14 ± 4.07 0.730 24.14 ± 4.30 23.93 ± 3.67 0.756 24.23 ± 4.18 23.84 ± 3.76 0.560 Education level (years) 0.612 0.505 0.184 ≤ 9 35 (25.2%) 18 (26.47) 17 (23.94) 17 (26.20) 18 (24.32) 21 (30.88) 14 (19.72) 10–15 52 (37.4%) 26 (38.24) 26 (36.62) 26 (40.00) 26 (35.14) 24 (35.29) 28 (39.44) ≥ 16 52 (37.4%) 24 (35.29) 28 (39.44) 22 (33.80) 30 (40.54) 23 (33.82) 29 (40.85) Ovarian reserve  FSH (IU/L) 6.40 (5.47, 7.86) 6.29 (5.35, 7.79) 6.45 (5.55, 7.91) 0.506 6.29 (5.51, 7.77) 6.42 (5.36, 7.88) 0.772 6.52 (5.51, 7.81) 6.25 (5.36, 7.87) 0.595  AMH (ng/mL) 2.74 (1.38, 4.38) 2.70 (1.34, 4.34) 2.76 (1.75, 4.68) 0.540 3.07 (2.19, 5.33) 2.53 (1.19, 4.01) 0.015 2.78 (1.37, 4.38) 2.69 (1.39, 4.38) 0.919 Trigger-day hormone levels  LH (IU/L) 3.15 (1.72, 6.30) 2.53 (1.54, 6.33) 3.34 (1.90, 5.67) 0.330 2.22 (1.40, 4.43) 4.09 (2.04, 6.85) 0.003 2.86 (1.46, 5.44) 3.41 (1.93, 6.44) 0.231  E2 (ng/L) 3525.00 (2058.00, 5850.00) 3181.50 (2028.75, 5733.00) 3831.00 (2058.00, 6003.00) 0.640 3744.00 (2219.50, 5812.50) 3249.00 (1994.00, 5930.25) 0.560 3505.50 (1862.25, 5975.25) 3693.00 (2065.00, 5775.00) 0.809  P (ng/mL) 3.03 (1.95, 4.61) 2.98 (1.88, 4.91) 3.03 (2.15, 4.00) 0.593 3.12 (1.79, 4.60) 2.61 (2.07, 4.63) 0.709 3.01 (1.77, 4.63) 3.03 (1.99, 4.60) 0.763 Method of Fertilization 0.352 0.664 0.300  IVF 105 (75.5%) 49 (72.06) 56 (78.87) 48 (73.85) 57 (77.03) 54 (79.41) 51 (71.83)  ICSI 34 (24.5%) 19 (27.94) 15 (21.13) 17 (26.15) 17 (22.97) 14 (20.59) 20 (28.17) BMI body mass index, FSH follicle-stimulating hormone, AMH anti-Müllerian hormone, LH luteinizing hormone, E 2 estradiol, P progesterone Differences between groups classified by ART outcomes were analyzed using the chi-square test for categorical variables and the Mann-Whitney U test for continuous variables Demographic characteristics of study participants by ART outcomes BMI body mass index, FSH follicle-stimulating hormone, AMH anti-Müllerian hormone, LH luteinizing hormone, E 2 estradiol, P progesterone Differences between groups classified by ART outcomes were analyzed using the chi-square test for categorical variables and the Mann-Whitney U test for continuous variables The detection rates for Zn, Fe, Mn, Ca, Mg, Cu, and Pb were 100.0%, while Cd was absent in all samples (this merely indicates that concentrations are under the detection threshold and does not eliminate the possibility of biological impacts). Among the detected metal elements, Fe had the highest median concentration at 76.76 µmol/L, while Mn had the lowest at 0.79 µmol/L (Table  2 ). Table 2 Distribution of trace elements concentration in follicular fluid Elements Detection rate (%) Median (P25, P75) Zn (µmol/L) 100.0 52.25 (41.97, 129.10) Fe (µmol/L) 100.0 76.76 (57.90, 91.57) Mn (µmol/L) 100.0 0.79 (0.59, 0.99) Ca (mmol/L) 100.0 1.96 (1.91, 2.06) Mg (mmol/L) 100.0 1.71 (1.53, 1.76) Cu (µmol/L) 100.0 27.23 (26.20, 28.13) Pb (µg/L) 100.0 12.62 (8.48, 16.67) Cd (µmol/L)) 0.0 / Distribution of trace elements concentration in follicular fluid Spearman correlation analysis revealed a notable positive correlation between Zn and Fe ( r  = 0.53, P  < 0.001), and between Ca and Mg ( r  = 0.41, P  < 0.001) (Fig.  1 ). Ca displayed significant negative correlations with both Fe ( r = -0.31, P  < 0.001) and Zn ( r = -0.18, P  = 0.033). No significant correlations were observed for Mn, Cu, or Pb with other metal elements analyzed (Fig.  1 ). Fig. 1 Spearman correlation matrix of the seven metal elements in follicular fluid ( n = 139). Color intensity represents the strength and direction of correlation coefficients (r), as shown in the scale bar. Asterisks denote statistical significance: * P  < 0.05, ** P  < 0.01, *** P  < 0.001 Spearman correlation matrix of the seven metal elements in follicular fluid ( n = 139). Color intensity represents the strength and direction of correlation coefficients (r), as shown in the scale bar. Asterisks denote statistical significance: * P  < 0.05, ** P  < 0.01, *** P  < 0.001 A correlation analysis was performed involving seven metal elements and the concentrations of FSH, AMH, LH, E2, and P (Table  3 ). Pb had a notable negative correlation with AMH ( r = -0.205, P  = 0.015) and E 2 ( r = -0.184, P  = 0.030). Zn demonstrated a positive correlation with LH levels ( r  = 0.231, P  = 0.006), while the Cu/Zn ratio was negatively correlated with LH ( r = -0.210, P  = 0.013). No significant relationships were observed for the other metals (Fe, Mn, Ca, Mg, Cu) with the hormone levels assessed. Table 3 Correlation between individual metal elements and hormone parameters Elements FSH (IU/L) AMH (ng/mL) LH (IU/L) E2 (ng/L) P (ng/mL) r P r P r P r P r P Zn (µmol/L) -0.123 0.150 -0.016 0.851 0.231 0.006 -0.065 0.450 -0.066 0.443 Fe (µmol/L) -0.156 0.066 0.080 0.350 0.160 0.060 0.092 0.280 0.066 0.437 Mn (µmol/L) 0.067 0.433 -0.059 0.491 -0.021 0.804 0.020 0.817 0.054 0.528 Ca (mmol/L) 0.032 0.709 -0.050 0.557 -0.057 0.504 -0.062 0.469 -0.087 0.306 Mg (mmol/L) -0.038 0.645 0.051 0.548 -0.027 0.751 0.009 0.914 < 0.001 0.999 Cu (µmol/L) 0.032 0.712 0.008 0.929 0.072 0.401 -0.034 0.689 0.001 0.994 Pb (µg/L) 0.032 0.709 -0.205 0.015 0.074 0.386 -0.184 0.030 -0.134 0.115 Cu/Zn 0.106 0.213 0.027 0.752 -0.210 0.013 0.072 0.398 0.077 0.371 FSH follicle-stimulating hormone, AMH anti-Müllerian hormone, LH luteinizing hormone, E 2 estradiol, P progesterone The Spearman correlation analysis was used Correlation between individual metal elements and hormone parameters FSH follicle-stimulating hormone, AMH anti-Müllerian hormone, LH luteinizing hormone, E 2 estradiol, P progesterone The Spearman correlation analysis was used As shown in Table  4 , the Ca concentration was markedly reduced in the high oocyte maturation rate group ( P  = 0.008), whereas Zn ( P  = 0.020) and Fe ( P  = 0.016) levels were significantly elevated in the higher high-quality embryo group. In contrast, no notable differences in metal element concentrations were found between groups with different clinical pregnancy outcomes. Table 4 Comparison of concentration of metal elements among the various ART outcomes groups Elements Oocyte maturation rate (%) P High-quality embryo rate (%) P Clinical pregnancy P < 87.5 ( n  = 68) ≥ 87.5 ( n  = 71) < 66.66 ( n  = 65) ≥ 66.66 ( n  = 74) No ( n  = 68) Yes ( n  = 71) Zn (µmol/L) 49.47 (40.26, 126.10) 53.96 (46.26, 130.00) 0.173 49.25 (39.04, 123.50) 54.16 (47.00, 130.48) 0.020 50.75 (41.01, 129.10) 53.96 (44.11, 130.00) 0.559 Fe (µmol/L) 69.12 (56.90, 88.02) 78.87 (61.94, 92.92) 0.070 68.46 (57.9, 84.87) 81.34 (58.91, 95.67) 0.016 77.50 (56.90, 92.42) 76.13 (59.25, 91.57) 0.943 Mn (µmol/L) 0.79 (0.59, 0.99) 0.69 (0.49, 0.99) 0.153 0.69 (0.54, 0.99) 0.79 (0.59, 0.99) 0.498 0.79 (0.59, 0.99) 0.69 (0.59, 0.99) 0.747 Ca (mmol/L) 1.99 (1.93, 2.11) 1.95 (1.86, 1.99) 0.008 1.96 (1.88, 2.11) 1.96 (1.92, 2.03) 0.713 1.96 (1.93, 2.08) 1.95 (1.88, 2.03) 0.223 Mg (mmol/L) 1.71 (1.57, 1.77) 1.71 (1.52, 1.76) 0.664 1.72 (1.52, 1.77) 1.71 (1.53, 1.76) 0.516 1.71 (1.54, 1.76) 1.71 (1.50, 1.77) 0.851 Cu (µmol/L) 27.36 (26.35, 27.96) 26.86 (25.81, 28.18) 0.294 26.97 (26.29, 27.84) 27.34 (26.01, 28.19) 0.514 13.37 (9.73, 17.05) 11.25 (8.06, 16.42) 0.159 Pb (µg/L) 11.98 (8.64, 17.15) 12.76 (8.44, 16.64) 0.985 11.90 (8.78, 15.32) 12.32 (8.46, 17.59) 0.310 27.00 (25.88, 28.05) 27.29 (26.31, 28.15) 0.595 Cu/Zn 0.55 (0.21, 0.67) 0.49 (0.21, 0.60) 0.107 0.55 (0.22, 0.70) 0.50 (0.21, 0.58) 0.034 0.53 (0.22, 0.64) 0.50 (0.21, 0.64) 0.632 Data was expressed as a median (P25, P75). Differences between groups were compared using the Mann-Whitney U test Comparison of concentration of metal elements among the various ART outcomes groups Data was expressed as a median (P25, P75). Differences between groups were compared using the Mann-Whitney U test The results of the binary logistic regression analyses are presented in Table  5 . In the primary analysis using metal concentration tertiles, several associations with raw P values < 0.05 were observed. Ca (high vs. low concentration: OR = 0.301, 95% CI: 0.126–0.718) and Cu (medium vs. low concentration: OR = 0.402, 95% CI: 0.171–0.949) levels were associated with lower oocyte maturation rates. Zn (high vs. low concentration: OR = 2.599, 95% CI: 1.096–6.164) and Fe ((high vs. low concentration: OR = 3.080, 95% CI: 1.277–7.431) levels were associated with higher high-quality embryo rates. Importantly, in sensitivity analyses using natural log-transformed continuous metal concentrations, the directions of the aforementioned associations were consistent with those observed in the primary tertile-based analysis (Table S2). After FDR adjustment, these associations were no longer statistically significant (Table  5 ), which may reflect limited statistical power due to sample size and multiple testing. Table 5 Associations between individual trace elements and ART outcomes: adjusted odds ratios with FDR correction Elements Oocyte maturation rate (%) P Q High-quality embryo rate (%) P Q Clinical pregnancy P Q ≥ 87.5 ( n  = 71) ≥ 66.66 ( n  = 74) Yes ( n  = 71) Zn (µmol/L)  low 1 - 1 - 1 -  medium 1.253 (0.542, 2.894) 0.598 0.914 1.940 (0.822, 4.576) 0.130 0.616 1.332 (0.567, 3.130) 0.511 0.914  high 1.845 (0.800, 4.254) 0.151 0.649 2.599 (1.096, 6.164) 0.030 0.264 1.530 (0.657, 3.560) 0.324 0.680 Fe (µmol/L)  low 1 - 1 - 1 -  medium 1.695 (0.739, 3.884) 0.213 0.649 1.182 (0.511, 2.734) 0.696 0.914 1.545 (0.665, 3.589) 0.312 0.649  high 1.925 (0.836, 4.432) 0.124 0.616 3.080 (1.277, 7.431) 0.012 0.210 1.072 (0.463, 2.483) 0.870 0.934 Mn (µmol/L)  low 1 - 1 - 1 -  medium 0.608 (0.263, 1.403) 0.244 0.649 1.113 (0.482, 2.572) 0.802 0.934 0.928 (0.399, 2.157) 0.862 0.934  high 0.586 (0.255, 1.350) 0.109 0.649 1.527 (0.655, 3.560) 0.327 0.649 0.803 (0.346, 1.864) 0.609 0.914 Ca (µmol/L)  low 1 - 1 - 1 -  medium 0.932 (0.395, 2.196) 0.871 0.934 1.171 (0.497, 2.760) 0.717 0.914 0.751 (0.319, 1.770) 0.513 0.914  high 0.301 (0.126, 0.718) 0.007 0.168 0.758 (0.327, 1.760) 0.519 0.914 0.617 (0.264, 1.439) 0.264 0.649 Mg (µmol/L)  low 1 - 1 - 1 -  medium 0.648 (0.282, 1.491) 0.308 0.649 0.831 (0.355, 1.942) 0.669 0.914 0.839 (0.361, 1.950) 0.683 0.914  high 0.760 (0.331, 1.748) 0.519 0.914 0.629 (0.270, 1.469) 0.284 0.649 1.216 (0.520, 2.841) 0.652 0.914 Cu (µmol/L)  low 1 - 1 - 1 -  medium 0.402 (0.171, 0.949) 0.038 0.264 0.404 (0.170, 0.959) 0.040 0.264 1.156 (0.494, 2.703) 0.739 0.914  high 0.647 (0.280, 1.491) 0.307 0.649 1.109 (0.471, 2.611) 0.812 0.914 1.078 (0.467, 2.490) 0.860 0.934 Pb (µg/L)  low 1 - 1 - 1 -  medium 0.946 (0.411, 2.175) 0.896 0.934 0.404 (0.170, 0.964) 0.041 0.264 0.299 (0.123, 0.726) 0.008 0.168  high 0.844 (0.364, 1.959) 0.693 0.914 0.893 (0.373, 2.135) 0.799 0.934 0.570 (0.238, 1.369) 0.208 0.649 Adjusted ORs (95% CI) from logistic regression models including age, BMI, AMH, E2, and stimulation protocol. Low concentration group as reference P values unadjusted, Q values adjusted using Benjamini-Hochberg FDR procedure Associations between individual trace elements and ART outcomes: adjusted odds ratios with FDR correction Adjusted ORs (95% CI) from logistic regression models including age, BMI, AMH, E2, and stimulation protocol. Low concentration group as reference P values unadjusted, Q values adjusted using Benjamini-Hochberg FDR procedure Stratified analyses by age (< 35 vs. ≥35 years) and BMI (< 24 vs. ≥24 kg/m²) revealed subgroup-specific associations (Table S1). In women aged < 35 years, Ca (high vs. low concentration: OR = 0.298, 95% CI: 0.125–0.708) was associated with lower oocyte maturation rates, while Zn (OR = 3.239, 95% CI: 1.186–8.850) and Fe (OR = 2.933, 95% CI: 1.098–7.840) were associated with high-quality embryo rates. Pb (medium vs. low concentration) showed negative associations with both high-quality embryo (OR = 0.253, 95% CI: 0.091–0.699) and clinical pregnancy rates (OR = 0.229, 95% CI: 0.081–0.645). In the BMI < 24 kg/m² subgroup, Zn (high vs. low concentration: OR = 4.494, 95% CI: 1.299–15.551) and Fe (high vs. low concentration: OR = 6.246, 95% CI: 1.718–22.709) were significantly linked to enhanced oocyte maturation rates, with Fe also associated with improved high-quality embryo rates. In contrast, Ca (high vs. low concentration: OR = 0.186, 95% CI: 0.054–0.641) was negatively correlated with oocyte maturation. Additionally, Fe (medium vs. low concentration: OR = 4.659, 95% CI: 1.355–16.022) was associated with higher clinical pregnancy rates, while Pb (medium vs. low concentration: OR = 0.226, 95% CI: 0.063–0.804) was linked to lower clinical pregnancy rates. In addition, Mn showed opposing effects: negative for oocyte maturation (OR = 0.238, 95% CI: 0.062–0.914) but positive for embryo quality (OR = 6.223, 95% CI: 1.351–28.674). Cu (medium vs. low concentration: OR = 0.227, 95% CI: 0.063–0.821) was negatively associated with high-quality embryo rates (Table S1). Restricted cubic spline analysis, adjusted for age, BMI, AMH, E 2 , and stimulation protocol revealed a significant linear positive association between Zn levels and high-quality embryo rates (P-overall = 0.045; P-nonlinear = 0.083). In contrast, Fe displayed a complex nonlinear relationship with high-quality embryo rates (P-overall = 0.007; P-nonlinear = 0.043), showing a negative correlation at lower concentrations that shifted to positive at higher levels (Fig.  2 A). Nevertheless, no significant statistical links were identified between any metal elements and the rates of oocyte maturation or clinical pregnancy outcomes (Figs.  2 B and C). Fig. 2 Restricted cubic splines (RCS) analysis for the associations of individual metal elements with ART outcomes ( n  = 139). Solid lines represent adjusted odds ratios (log scale) with 95% confidence intervals (shaded areas). Horizontal dashed lines indicate no effect (odds ratio = 1). All models were adjusted for age, BMI, AMH, E2, and stimulation protocol. Additionally, four knots were incorporated at the 5th, 35th, 65th, and 95th percentiles of the metal concentration distributions. Metal concentrations were standardized as z-scores. A High-quality embryo rate. B Oocyte maturation rate. C Clinical pregnancy Restricted cubic splines (RCS) analysis for the associations of individual metal elements with ART outcomes ( n  = 139). Solid lines represent adjusted odds ratios (log scale) with 95% confidence intervals (shaded areas). Horizontal dashed lines indicate no effect (odds ratio = 1). All models were adjusted for age, BMI, AMH, E2, and stimulation protocol. Additionally, four knots were incorporated at the 5th, 35th, 65th, and 95th percentiles of the metal concentration distributions. Metal concentrations were standardized as z-scores. A High-quality embryo rate. B Oocyte maturation rate. C Clinical pregnancy BKMR analysis, adjusted for age, BMI, AMH, E 2 , and stimulation protocol revealed distinct mixture and individual metal effects on ART outcomes. In the BKMR model, the posterior inclusion probability (PIP) reflects the importance of each metal, with a value closer to 1 indicating higher importance. In variable selection, Fe showed the highest importance for high-quality embryo rates (PIP = 0.808), exhibiting a nonlinear relationship that shifted from negative to positive with increasing concentrations (Fig.  3 A). In terms of oocyte maturation, Ca showed the strongest association (PIP = 0.676), with levels above the median significantly negatively correlated with maturation rates (Fig.  3 D). Fig. 3 Bayesian Kernel Machine Regression (BKMR) analysis of metal mixtures and ART outcomes ( n  = 139). Solid lines and data points represent estimated effect sizes, with shaded areas indicating 95% credible intervals. Models were adjusted for age, BMI, AMH, E2, and stimulation protocol. Metal concentrations were standardized as z-scores. A Individual metal effects on high-quality embryo rate. B Cumulative mixture effects on high-quality embryo rate. C Single-metal effect (75th vs. 25th quantile) on high-quality embryo rate, with other metals fixed at the 25th, 50th, or 75th quantile. D Individual effects of single metals on oocyte maturation rate. E Cumulative mixture effects on oocyte maturation rate. F Single-metal effect (75th vs. 25th quantile) on oocyte maturation rate, with other metals fixed at the 25th, 50th, or 75th quantile Bayesian Kernel Machine Regression (BKMR) analysis of metal mixtures and ART outcomes ( n  = 139). Solid lines and data points represent estimated effect sizes, with shaded areas indicating 95% credible intervals. Models were adjusted for age, BMI, AMH, E2, and stimulation protocol. Metal concentrations were standardized as z-scores. A Individual metal effects on high-quality embryo rate. B Cumulative mixture effects on high-quality embryo rate. C Single-metal effect (75th vs. 25th quantile) on high-quality embryo rate, with other metals fixed at the 25th, 50th, or 75th quantile. D Individual effects of single metals on oocyte maturation rate. E Cumulative mixture effects on oocyte maturation rate. F Single-metal effect (75th vs. 25th quantile) on oocyte maturation rate, with other metals fixed at the 25th, 50th, or 75th quantile The overall analysis of mixture effects revealed that as the concentrations of all metals shifted from the 50th to the 25th-75th percentile range (in 0.05 increments), an increase in metal mixture exposure correlated with a rise in high-quality embryo rates but a decline in oocyte maturation rates (Fig.  3 B and E). Single-metal dose-response evaluations, which involved raising the concentration of the target metal from the 25th to the 75th percentile while keeping the other six metals at their respective 25th, 50th, or 75th percentiles, did not show significant effects for individual metals (Fig.  3 C and F). Sensitivity analyses excluding extreme values (top and bottom 2.5%), with models adjusted for age and BMI, confirmed the robust associations of Fe with embryo quality and Ca with oocyte maturation (Figure S1), with significant positive (Fe) and negative (Ca) single-metal effects observed (Figures S1C, S1F).

This study underscores the significance of metal elements in FF as biomarkers, which can effectively reflect the environmental exposure levels of oocytes and embryo development. Our results show strong correlations between these metal elements and important parameters in ART. Notably, higher levels of Zn and Fe were positively associated with high-quality embryo rates, whereas elevated level of Ca was negatively correlated with oocyte maturation. Additionally, Pb exposure was identified as a risk factor associated with lower rates of high-quality embryos and unfavorable clinical pregnancy outcomes. These findings underscore the critical importance of considering metal exposure in relation to female fertility. Further research is needed to elucidate how FF metals mechanistically affect female fertility.

FF, the microenvironment nurturing the developing oocyte, is a critical medium for assessing direct metal exposure relevant to ART outcomes [ 9 ]. Research shows that the metal levels found in larger follicles provide a more accurate representation of recent blood exposure [ 11 ]. This study characterized the spectrum of metal exposure in FF from a cohort of reproductive-aged women in the Jiaodong Peninsula area of China and evaluated its association with key ART parameters. Our findings indicated that specific metals exert differential effects on embryo quality, oocyte maturation, and pregnancy outcomes, with mechanisms potentially related to oxidative stress (Fig. 4 ). To our knowledge, this is one of the largest studies investigating metal exposure profiles in FF from this region. Fig. 4 Detection of metals in follicular fluid. Analysis of eight metal elements in follicular fluid revealed that Pb, excess Cu, and Ca impair oocyte quality and embryonic development by inducing oxidative stress, whereas Zn and Fe exert protective effects via the antioxidant system. Antagonistic (Ca-Zn, Ca-Fe) and synergistic (Zn-Fe) interactions were observed among elements. Pb showed negative correlations with AMH and E₂, while Zn positively correlated with LH Detection of metals in follicular fluid. Analysis of eight metal elements in follicular fluid revealed that Pb, excess Cu, and Ca impair oocyte quality and embryonic development by inducing oxidative stress, whereas Zn and Fe exert protective effects via the antioxidant system. Antagonistic (Ca-Zn, Ca-Fe) and synergistic (Zn-Fe) interactions were observed among elements. Pb showed negative correlations with AMH and E₂, while Zn positively correlated with LH Pb is recognized as one of the primary heavy metal contaminants that significantly endangers human health. Ideally, the concentration of Pb in the human body should be nonexistent [ 12 ]. Our study identified strong negative relationships between Pb levels in FF and both the rates of high-quality embryos and successful clinical pregnancies, aligning with findings from various earlier studies [ 11 , 13 ]. During the embryonic stage, studies have demonstrated a negative correlation between serum Pb concentration and the rate of high-quality embryos, which aligns with our research findings [ 8 ]. Another investigation found a positive relationship between hair Pb levels and the likelihood of mature oocytes, potentially due to low-level Pb’s stimulating effects via estrogen receptor modulation [ 14 ]. These detrimental effects likely arise from Pb’s multifaceted toxicity, which disrupts metabolism, impairs placental function, and induces oxidative stress and apoptosis [ 15 – 17 ]. Notably, our study demonstrated more pronounced negative effects of Pb in the population aged < 35 years with BMI < 24 kg/m². We speculate that endocrine changes associated with elevated BMI may buffer Pb toxicity. Consequently, it is crucial for clinical practices to highlight the reproductive hazards associated with environmental Pb exposure. Zn, the important trace element in the human body, is a vital cofactor for more than 300 enzymes and is crucial in the processes of oocyte meiosis and follicular rupture [ 18 , 19 ]. Our study uncovered a strong positive link between Zn levels in FF and the rates of high-quality embryos, aligning with findings from previous research that indicated a higher risk of IVF-ET failure in individuals with serum Zn levels below 610.3 ng/mL [ 20 ]. Zn aids embryonic development by activating antioxidant enzyme systems and modulating the functions of cumulus cells [ 21 ]. On the other hand, Zn deficiency can disrupt oocyte epigenetic reprogramming and maintain transcriptomic balance [ 22 ]. Thus, it is essential to balance dietary Zn intake while controlling environmental exposure. Disruption of ovarian Fe homeostasis can lead to the excessive production of reactive oxygen species, resulting in oxidative stress disorders and decreased ovarian function [ 23 ]. Our study found a significant positive relationship between Fe levels in FF and high-quality embryo rates. It appeared that Fe might play a crucial role in the successful in vitro culture of 8-cell embryos and blastocysts. Prolonged Fe deficiency increases the number of apoptotic cleavage cells, which aligns with the findings of our research [ 24 ]. However, excessive Fe adversely affects oocyte maturation and blastocyst formation via the Fe death pathway. Research indicates that patients with endometriosis exhibit significantly elevated Fe concentrations in FF. This increase induces the downregulation of GPX4 protein and enhances lipid peroxidation, ultimately impairing oocyte maturation and reducing fertilization rates [ 25 ]. Consistently, our study found that Fe has concentration-dependent dual effects in the follicular microenvironment, with the inflection point in RCS curves possibly indicating a critical threshold for Fe’s biological impact. Consequently, future studies should incorporate multicenter cohorts to accurately determine the optimal threshold for enhancing Fe homeostasis. Ca²⁺, as vital secondary messengers, plays an essential role in regulating oocyte maturation and fertilization [ 26 ]. Our study indicated a significant negative relationship between Ca levels in FF and the rate of oocyte maturation, a trend that persisted in groups of women aged under 35 and with a BMI below 24 kg/m², which is consistent with previous reports [ 4 ]. Ca²⁺ overload may inhibit oocyte maturation by impairing mitochondrial function, inducing oxidative stress, and delaying meiotic recovery [ 27 ]. Long-term exposure to high-Ca environments, such as regions with elevated water hardness or industrial dust pollution, may exacerbate the imbalance of Ca metabolism in oocytes, thereby adversely affecting female fertility [ 28 , 29 ]. Additionally, the overuse of Cu-based medications, dietary supplements, or intrauterine devices can result in Cu toxicity [ 30 ]. Our study found a negative correlation between Cu levels in FF and both oocyte maturation and high-quality embryo rates, supporting earlier findings that prolonged Cu exposure negatively affects follicular and embryonic development [ 13 ]. Excessive Cu can induce ovarian ultrastructural damage through mechanisms such as apoptosis, cytoplasmic vacuolization, and disruption of cell membrane structure [ 31 , 32 ]. Notably, our study revealed a negative correlation between the FF Cu-Zn ratio and the rates of high-quality embryos. Conversely, some studies suggest that a higher Cu-Zn ratio is associated with better ovarian response, embryo quality and increased oocyte yield [ 33 , 34 ]. The reasons for this discrepancy remain unclear but may involve differences in population characteristics, exposure ranges, or the interplay between absolute concentrations and their ratio. Metal elements engage in interactions that subsequently affect reproductive outcomes via endocrine mechanisms. In the present study, we found notable negative correlations between Ca and both Zn and Fe. The Ca 2+ reaction and IP3R1 function in eggs necessitate a specific Zn 2+ concentration range to ensure optimal responses to fertilization and egg activation [ 35 ]. Additionally, Ca can inhibit DMT1 in a dose-dependent manner, thereby reducing Fe absorption, which is consistent with our results [ 36 ]. Our study identified a significant positive relationship between Ca and Mg. Ca 2+ and Mg 2+ are classified as hard Lewis acids, and they exhibit a preference for binding with hard Lewis bases [ 37 ]. Consequently, both ions demonstrate moderate binding strength to organic ligands. Furthermore, Ca 2+ signaling plays a crucial role in the mechanisms underlying cell survival and apoptosis, necessitating Mg 2+ as an essential cofactor for various enzymes that regulate ATP-related cellular activities. Additionally, Mg 2+ serves as a regulatory factor for Ca 2+ ion channels, potentially counteracting the effects of Ca [ 37 ]. These reports provide robust support for our research findings. Our study also uncovered significant negative associations between Pb levels in FF and both AMH and E₂ concentrations. Notably, a correlation with E₂ was observed, suggesting that the disruption of estrogenic signaling may represent a particularly sensitive pathway of Pb toxicity in the ovarian follicle. It reinforces the idea that Pb may impair reproductive function by diminishing E₂ receptor affinity and disrupting E₂-mediated signaling pathways [ 38 ]. Recent research has identified a negative correlation between blood Pb concentration and ovarian reserve function, thereby providing substantial support for our study [ 39 ]. Furthermore, we found that positive correlations were observed between Zn and LH levels, while the ratio of Cu to Zn exhibited negative correlations with LH. Pascua et al. showed that LH significantly influences Zn distribution in oocytes and cumulus cells, thereby affecting oocyte maturation through the regulation of Zn transporters [ 40 ]. The research findings presented here align closely with our study. While the specific mechanisms underlying these combined effects remain unclear, our results offer novel insights into the enhancement of female reproductive capabilities. This study presents several limitations. Firstly, as the evaluation of FF metals and outcomes occurred within a single ART cycle, the observed associations should not be interpreted as causal effects. Secondly, our modest sample size may have limited statistical power, a constraint further exacerbated by the stringent Benjamini-Hochberg FDR correction applied for multiple comparisons. Thirdly, although models were adjusted for key covariates (age, BMI, AMH, estradiol, and stimulation protocol), we lacked data on several important confounders due to data availability constraints, including smoking, alcohol use, socioeconomic factors, dietary habits, and the hormonal microenvironment of individual follicles. The research framework did not consider the cumulative effects of multiple embryo transfers or metal exposure data from male partners, both of which could significantly affect reproductive results. Finally, the focus on couples undergoing ART limits the applicability of the findings to the broader infertile population. Nonetheless, considering the widespread exposure to metal elements and the similar mechanisms of reproductive toxicity, our results are still important for understanding how environmental factors affect reproductive health. Future research should aim for larger sample sizes, utilize longitudinal approaches, and include environmental exposure data from both partners to achieve a more thorough evaluation of the combined effects of metal elements on reproductive health.

Amid rising global environmental contamination and public health challenges, infertility among women aged 15–49 has increased steadily, with epidemiological forecasts indicating that this pattern will continue through 2040 [ 1 ]. The use of assisted reproductive technology (ART), including in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI), has seen a notable rise [ 2 ]. As industrialization advances, the levels of metal pollutants emitted into the air, water, and soil are escalating [ 3 ]. Consequently, elucidating the dose-response relationship between exposure to these metals and critical parameters of ART is of clinical significance and may provide new insights for evaluating risks to reproductive health. The human body contains essential metal elements, which are primarily acquired through food, water, and environmental exposure. Heavy metals can enter the human body via the digestive system, lungs, and skin. Research indicates that these metals may pose a significant threat to reproductive health [ 4 ]. For instance, lead (Pb) and cadmium (Cd) exert profound reproductive toxicity through inducing oxidative stress, disrupting the hypothalamic-pituitary-gonadal axis, and triggering cell apoptosis, thereby impairing reproductive function [ 5 ]. These heavy metals accumulate in reproductive tissues, interfere with hormone synthesis, and cause DNA damage, posing significant threats to fertility and reproductive health [ 6 ]. Although essential elements are present in trace amounts, they serve as cofactors for over 300 enzymes, playing a crucial role in maintaining normal reproductive function [ 7 ]. Elements like zinc (Zn), iron (Fe), manganese (Mn), calcium (Ca), magnesium (Mg), and copper (Cu) play significant roles in biological functions, especially in cellular metabolism and antioxidant protection, which are essential for the development potential of oocytes. Current investigations mainly concentrate on blood and urine samples [ 8 ], while research on follicular fluid (FF), an important reproductive environment, remains scarce. FF plays a crucial role within ovarian follicles, acting as a medium for communication between oocytes and providing a supportive environment for their growth [ 2 ]. It is the first location where oocytes come into contact with external environment and alterations in this fluid can influence oocyte maturation, fertilization, and the early stages of embryonic development [ 9 ]. Compared to blood and urine, FF serves as a microenvironment for oocyte development, facilitating direct interaction with the oocyte and providing a more accurate reflection of metal exposure. This study analyzes the concentrations of eight metallic elements in the FF of women of reproductive age participating in ART cycles in the Jiaodong Peninsula area of China. The eight elements include six that are essential and two that are non-essential. We employed logistic regression, restricted cubic splines (RCS), and Bayesian kernel machine regression (BKMR) methods to explore the relationships between these metal exposures and key outcome measures (oocyte maturation rate, high-quality embryo rate, and clinical pregnancy). This study aims to provide novel insights and strategies for epidemiological and experimental research concerning the impact of environmental factors on female reproductive health.

Supplementary Material 1. Supplementary Material 1.