Ovarian volume partially explains associations of phthalate biomarkers with anti-Müllerian hormone and estradiol in midlife women

Abstract

Background/objectives: Women are ubiquitously exposed to endocrine disruptors, including phthalates. Ovarian follicles undergoing folliculogenesis (indirectly measured by ovarian volume) produce anti-Müllerian hormone (AMH) and estradiol (E2). We evaluated associations of phthalates with ovarian volume to assess whether this explained prior positive associations of phthalates with AMH and E2.

Methods: Women ages 45-54 years (n = 614) had transvaginal ultrasounds of right/left ovaries to calculate mean ovarian volume. Women provided up-to-four urine and blood samples for quantifying AMH (first serum sample), E2 (all serum samples), and nine phthalate metabolites (from pooled urine, representing six parent phthalates). Multivariable linear or logistic regression models (for individual phthalate biomarkers), as well as weighted quantile sum (WQS) regression (for mixture analyses) evaluated associations of phthalate biomarkers with ovarian volume. Using cross-sectional mediation analysis, we assessed whether associations of phthalates with ovarian volume partially explained those of phthalates with AMH or E2.

Results: Most women were non-Hispanic White (68%) and pre-menopausal (67%) with higher urinary phthalate metabolite concentrations than U.S. women. In single-pollutant models, 10% increases in mono(3-carboxypropyl) phthalate (MCPP) and monobenzyl phthalate (MBzP) were associated with 0.44% (95% CI: -0.02%, 0.91%) and 0.62% (95% CI: 0.02%, 1.23%) larger ovarian volumes, respectively. As a cumulative mixture, 10% increases in the phthalate mixture were associated with 2.89% larger ovarian volume (95%CI: 0.27, 5.59) with MCPP (35%) and MBzP (41%) identified as major contributors. Higher ovarian volume due to a 10% increase in MBzP (indirect effect OR: 1.004; 95% CI: 1.00, 1.01) explained 16% of the positive association between MBzP and higher AMH, whereas higher ovarian volume due to a 10% increase in MCPP (indirect effect %Δ: 0.11; 95% CI: -0.01, 0.22) explained 23% of the positive association between MCPP and E2.

Conclusion: In this cross-sectional study, phthalates were associated with increased ovarian volume, with implications for midlife hormone production.

Keywords: Anti-Müllerian hormone; Estradiol; Midlife; Ovarian volume; Phthalates.

Fig. 1.. Relative importance of phthalate biomarkers in associations between the biomarker mixture and ovarian volume.

The length of the blue (left) and pink (right) bars represent the relative weights (as percentages) of each biomarker, which were generated from negatively- (blue) and positively (pink)-constrained WQS regression models with repeated holdouts. Weights above 16.7% were identified as notable contributors to meaningful cumulative associations. Models accounted for race/ethnicity, income, marital status, smoking status, age, oral contraceptive use, change in BMI since age 18, alcohol use, depression symptoms, and reproductive history. ∑DEHP, sum of di(2-ethylhexyl) phthalate metabolites; MBP, monobutyl phthalate; MBzP, monobenzyl phthalate; MCPP, mono(3-carboxypropyl) phthalate; MEP, monoethyl phthalate; MiBP, monoisobutyl phthalate; WQS, Weighted Quantile Sum regression.

Fig. 2.. Associations of MCPP and MBzP with AMH and E2 mediated by ovarian volume.

Shaded arrows/shapes indicate meaningful associations for the direct effects (MCPP or MBzP → AMH or E2) and indirect effects (MCPP or MBzP → ovarian volume → AMH or E2), whereas unshaded arrows/shapes indicate null findings. Percentage mediated indicates the proportion of the total relationship that is explained by the indirect effect; crossed circle indicated no meaningful percentage mediated. Formal mediation analysis evaluated the total, natural direct, and natural indirect effects accounting for race/ethnicity, income, marital status, smoking status, age, oral contraceptive use, change in BMI since age 18, alcohol use, depression symptoms, and reproductive history. AMH, anti-Müllerian hormone; E2, estradiol; MBzP, monobenzyl phthalate; MCPP, mono(3-carboxypropyl) phthalate.