Serum and follicular fluid metabolome and markers of ovarian stimulation
Robert B Hood,Donghai Liang,Youran Tan,Jennifer B Ford,Irene Souter,Jorge E Chavarro,Dean P Jones,Russ Hauser,Audrey J Gaskins
DOI: https://doi.org/10.1093/humrep/dead189
2023-11-02
Abstract:Study question: What metabolic pathways and metabolites in the serum and follicular fluid are associated with peak estradiol levels and the number of mature oocytes? Summary answer: In the serum metabolome, mostly fatty acid and amino acid pathways were associated with estradiol levels and mature oocytes while in the follicular fluid metabolome, mostly lipid, vitamin, and hormone pathways were associated with peak estradiol levels and mature oocytes. What is known already: Metabolomics has identified several metabolic pathways and metabolites associated with infertility but limited data are available for ovarian stimulation outcomes. Study design, size, duration: A prospective cohort study of women undergoing IVF from 2009 to 2015. Participants/materials, setting, methods: A total of 125 women undergoing a fresh IVF cycle at a fertility clinic in the Northeast United States who provided a serum and follicular fluid sample. Untargeted metabolomics profiling was conducted using liquid chromatography with high-resolution mass spectrometry in two chromatography columns (C18 and hydrophilic interaction chromatography (HILIC)). The main ovarian stimulation outcomes were peak serum estradiol levels and number of mature oocytes. We utilized adjusted generalized linear regression models to identify significant metabolic features. Models were adjusted for age,BMI, initial infertility diagnosis, and ovarian stimulation protocol. We then conducted pathway analysis using mummichog and metabolite annotation using level-1 evidence. Main results and role of chance: In the serum metabolome, 480 and 850 features were associated with peak estradiol levels in the C18 and HILIC columns, respectively. Additionally, 437 and 538 features were associated with mature oocytes in the C18 and HILIC columns, respectively. In the follicular fluid metabolome, 752 and 929 features were associated with peak estradiol levels in the C18 and HILIC columns, respectively, Additionally, 993 and 986 features were associated with mature oocytes in the C18 and HILIC columns, respectively. The most common pathways associated with peak estradiol included fatty acids (serum and follicular fluid), hormone (follicular fluid), and lipid pathways (follicular fluid). The most common pathways associated with the number of mature oocytes retrieved included amino acids (serum), fatty acids (serum and follicular fluid), hormone (follicular fluid), and vitamin pathways(follicular fluid). The vitamin D3 pathway had the strongest association with both ovarian stimulation outcomes in the follicularfluid. Four and nine metabolites were identified using level-1 evidence (validated identification) in the serum and follicular fluid metabolomes, respectively. Limitations, reasons for caution: Our sample was majority White and highly educated and may not be generalizable to thewider population. Additionally, residual confounding is possible and the flushing medium used in the follicular fluid could have diluted our results. Wider implications of the findings: The pathways and metabolites identified by our study provide novel insights into the biologicalmechanisms in the serum and follicular fluid that may underlie follicular and oocyte development, which could potentially be used to improve ovarian stimulation outcomes. Study funding/competing interest(s): This work was supported by the following grants from the National Institute of Environmental Health Sciences (P30-ES019776, R01-ES009718, R01-ES022955, P30-ES000002, R00-ES026648, and T32-ES012870), and National Institute of Diabetes and Digestive and Kidney Diseases (P30DK046200). The authors have no competing interests to disclose. Trial registration number: N/A.