ABSTRACT
Although prior studies have found associations of the ovarian reserve with urinary concentrations of some individual phenols and phthalate metabolites, little is known about the potential associations of these chemicals as a mixture with the ovarian reserve. We investigated whether mixtures of four urinary phenols (bisphenol A, butylparaben, methylparaben, propylparaben) and eight metabolites of five phthalate diesters including di(2-ethylhexyl) phthalate were associated with markers of the ovarian reserve among 271 women attending a fertility center who enrolled in the Environment and Reproductive Health study (2004-2017). The analysis was restricted to one outcome per study participant using the earliest outcome after the last exposure assessment. Ovarian reserve markers included lower antral follicle count (AFC) defined as AFC < 7, circulating serum levels of day 3 follicle stimulating hormone (FSH) assessed by immunoassays, and diminished ovarian reserve (DOR) defined as either AFC < 7, FSH > 10 UI/L or primary infertility diagnosis of DOR. We applied Bayesian Kernel Machine Regression (BKMR) and quantile g-computation to estimate the joint associations and assess the interactions between chemical exposure biomarkers on the markers of the ovarian reserve while adjusting for confounders. Among all 271 women, 738 urine samples were collected. In quantile g-computation models, a quartile increase in the exposure biomarkers mixture was not significantly associated with lower AFC (OR = 1.10, 95 % CI = 0.52, 2.30), day 3 FSH levels (Beta = 0.30, 95 % CI = -0.32, 0.93) or DOR (OR = 1.02, 95 % CI = 0.52, 2.05). Similarly, BKMR did not show any evidence of associations between the mixture and any of the studied outcomes, or interactions between chemicals. Despite the lack of associations, these results need to be explored among women in other study cohorts.
Subject(s)
Diethylhexyl Phthalate , Infertility, Female , Ovarian Reserve , Humans , Female , Fertility Clinics , Bayes Theorem , Infertility, Female/diagnosis , Infertility, Female/urine , Follicle Stimulating Hormone , BiomarkersABSTRACT
RESEARCH QUESTION: Do heavy metals affect the risk of diminished ovarian reserve (DOR) in women of reproductive age? DESIGN: A total of 139 cases and 153 controls were included between 2016 and 2020. The participants were aged between 18 and 40 years and attended consultations for couple infertility in one of four fertility centres in western France. Cases of DOR were defined as women with an antral follicle count less than 7, anti-Müllerian hormone levels 1.1 ng/ml or less, or both. Controls were frequency matched on age groups and centres, and were women with normal ovarian reserve evaluations, no malformations and menstrual cycles between 26 and 35 days. Heavy metals (lead, mercury, cadmium and chromium) were measured in whole blood at inclusion. Single-exposure associations were examined with multivariable logistic regressions adjusted on potential confounders. Mixture effects were investigated with quantile g-computation and Bayesian kernel machine regression (BKMR). RESULTS: Chromium as a continuous exposure was significantly associated with DOR in unadjusted models (OR 2.07, 95% CI 1.04 to 4.13) but the association was no longer significant when confounders were controlled for (adjusted OR 2.75, 95% CI 0.88 to 8.60). Similarly, a statistically significant association was observed for the unadjusted second tercile of cadmium exposure (OR 1.87, 95% CI 1.06 to 3.30); however, this association was no longer statistically significant after adjustment. None of the other associations tested were statistically significant. Quantile g-computation and BKMR both yielded no significant change of risk of DOR for the mixture of metals, with no evidence of interaction. CONCLUSIONS: Weak signals that some heavy metals could be associated with DOR were detected. These findings should be replicated in other studies.
Subject(s)
Metals, Heavy , Ovarian Diseases , Ovarian Reserve , Humans , Female , Adolescent , Young Adult , Adult , Infant, Newborn , Male , Cadmium , Bayes Theorem , Chromium , Anti-Mullerian HormoneABSTRACT
The associations between urinary phenol concentrations and markers of thyroid function and autoimmunity among potentially susceptible subgroups, such as subfertile women, have been understudied, especially when considering chemical mixtures. We evaluated cross-sectional associations of urinary phenol concentrations, individually and as a mixture, with serum markers of thyroid function and autoimmunity. We included 339 women attending a fertility center who provided one spot urine and one blood sample at enrollment (2009-2015). We quantified four phenols in urine using isotope dilution high-performance liquid chromatography-tandem mass spectrometry, and biomarkers of thyroid function (thyroid-stimulating hormone (TSH), free and total thyroxine (fT4, TT4), and triiodothyronine (fT3, TT3)), and autoimmunity (thyroid peroxidase (TPO) and thyroglobulin (Tg) antibodies (Ab)) in serum using electrochemoluminescence assays. We fit linear and additive models to investigate the association between urinary phenols-both individually and as a mixture-and serum thyroid function and autoimmunity, adjusted for confounders. As a sensitivity analysis, we also applied Bayesian Kernel Machine Regression (BKMR) to investigate non-linear and non-additive interactions. Urinary bisphenol A was associated with thyroid function, in particular, fT3 (mean difference for a 1 log unit increase in concentration: -0.088; 95% CI [-0.151, -0.025]) and TT3 (-0.066; 95% CI [-0.112, -0.020]). Urinary methylparaben and triclosan were also associated with several thyroid hormones. The overall mixture was negatively associated with serum fT3 concentrations (mean difference comparing all four mixture components at their 75th vs. 25th percentiles: -0.19, 95% CI [-0.35, -0.03]). We found no evidence of non-linearity or interactions. These results add to the current literature on phenol exposures and thyroid function in women, suggesting that some phenols may alter the thyroid system.
ABSTRACT
Exposure to per- and polyfluoroalkyl substances (PFAS) through the environment can lead to harmful health outcomes and the development of disease. However, little is known about how PFAS impact underlying biology that contributes to these adverse health effects. The metabolome represents the end product of cellular processes and has been used previously to understand physiological changes that lead to disease. In this study, we investigated whether exposure to PFAS was associated with the global, untargeted metabolome. In a cohort of 459 pregnant mothers and 401 children, we quantified plasma concentrations of six individual PFAS- PFOA, PFOS, PFHXS, PFDEA, and PFNA- and performed plasma metabolomic profiling by UPLC-MS. In adjusted linear regression analysis, we found associations between plasma PFAS and perturbations in lipid and amino acid metabolites in both mothers and children. In mothers, metabolites of 19 lipid pathways and 8 amino acid pathways were significantly associated with PFAS exposure at an FDR<0.05 threshold; in children, metabolites of 28 lipid pathways and 10 amino acid pathways exhibited significant associations at FDR<0.05 with PFAS exposure. Our investigation found that metabolites of the Sphingomyelin, Lysophospholipid, Long Chain Polyunsaturated Fatty Acid (n3 and n6), Fatty Acid- Dicarboxylate, and Urea Cycle showed the most significant associations with PFAS, suggesting these may be particular pathways of interest in the physiological response to PFAS. To our knowledge, this is the first study to characterize associations between the global metabolome and PFAS across multiple periods in the life course to understand impacts on underlying biology, and the findings presented here are relevant in understanding how PFAS disrupt normal biological function and may ultimately give rise to harmful health effects.