Antioxidants linked to embryo counts, not pregnancy outcomes in DOR IVF study

In a retrospective study of 949 women with diminished ovarian reserve undergoing IVF or intracytoplasmic sperm injection, coenzyme Q10 use was associated with a higher number of transferable embryos, while combined coenzyme Q10 and vitamin E use was associated with a higher number of top-quality embryos. Neither association translated into statistically significant differences in clinical pregnancy, live birth or early miscarriage rates. The investigators at the First Affiliated Hospital of Anhui Medical University in Hefei, China, described the findings as exploratory laboratory signals that require confirmation rather than evidence that the supplements should routinely be added to fertility treatment.

The study examined the first IVF/ICSI-embryo transfer cycle for each participant treated at the center between January 2018 and December 2023. Diminished ovarian reserve was diagnosed when the total antral follicle count across both ovaries was five or fewer and at least one additional criterion was met: an anti-Müllerian hormone concentration of no more than 1.2 ng/mL or a basal follicle-stimulating hormone concentration of at least 10 IU/L. Participants were assigned to four groups according to documented supplement use before controlled ovarian stimulation: 338 used neither supplement, 260 used both vitamin E and coenzyme Q10, 224 used coenzyme Q10 alone and 127 used vitamin E alone.

Treatment was not randomly assigned. Physicians recommended the supplements at their discretion, and patients received them only if they accepted the recommendation. Coenzyme Q10 was administered at 200 mg three times daily for two months, and vitamin E at 100 mg three times daily for two months before ovarian stimulation. The groups did not differ significantly in measured baseline characteristics, including patient and partner age, body mass index, infertility duration, primary infertility status, basal FSH, antral follicle count, AMH and estradiol. Stimulation protocols, total gonadotropin doses, stimulation duration, IVF or ICSI use, oocyte retrieval, mature oocytes, two-pronuclear embryos, transfer day, number of embryos transferred and the proportion of fresh transfers also showed no significant overall differences.

The clearest unadjusted differences appeared in embryo counts. The combined-supplement group had more transferable embryos than the control and vitamin E-only groups, while coenzyme Q10 alone also had more transferable embryos than those two groups. The difference between combined treatment and coenzyme Q10 alone was not statistically significant. For top-quality embryos, the combination had higher counts than the control, coenzyme Q10-only and vitamin E-only groups, with all three comparisons producing P values below 0.001. Cycle-cancellation rates were not significantly different among the four documented treatment categories: 0.59% in controls, 2.99% with both supplements, 2.18% with coenzyme Q10 alone and 3.05% with vitamin E alone, based on 968 otherwise eligible cycles.

The primary statistical models were multivariable Poisson regressions with a log link and Huber–White robust standard errors. They accounted for patient and partner age, body mass index, infertility duration, primary infertility, basal FSH, antral follicle count, AMH, estradiol, use of a gonadotropin-releasing hormone antagonist protocol, total gonadotropin dose, stimulation duration, fertilization method and calendar year of cycle initiation. Compared with no supplementation, combined treatment was associated with 21.4% more transferable embryos, with an incidence rate ratio of 1.214 and a 95% confidence interval of 1.076–1.368. Coenzyme Q10 alone was associated with 14.8% more, with an incidence rate ratio of 1.148 and a 95% confidence interval of 1.022–1.289. Vitamin E alone showed no statistically significant association with transferable-embryo count.

A separate 2-by-2 model treated vitamin E use, coenzyme Q10 use and their interaction as binary factors. Coenzyme Q10 remained associated with transferable-embryo count, whereas neither vitamin E nor the interaction term was statistically significant for that outcome. This pattern suggests that the observed association with additional transferable embryos was driven predominantly by coenzyme Q10 use, although the retrospective design cannot establish causation. Adjusted model-based averages were 0.910 transferable embryos for controls, 1.104 for combined treatment, 1.045 for coenzyme Q10 alone and 0.882 for vitamin E alone, with respective 95% confidence intervals of 0.830–0.990, 1.013–1.196, 0.966–1.123 and 0.777–0.986.

Top-quality embryos followed a different pattern. Combined supplementation was associated with 57.8% more top-quality embryos than control treatment, 85.8% more than coenzyme Q10 alone and 115.8% more than vitamin E alone. The corresponding adjusted incidence rate ratios were 1.578, 1.858 and 2.158, with 95% confidence intervals of 1.307–1.906, 1.483–2.328 and 1.638–2.843. The vitamin E-by-coenzyme Q10 interaction was positive on the multiplicative scale, with an incidence rate ratio of 2.541 and a 95% confidence interval of 1.767–3.654. Adjusted averages were 0.478 top-quality embryos for controls, 0.754 for combined treatment, 0.406 for coenzyme Q10 alone and 0.349 for vitamin E alone, each with its own 95% confidence interval.

The interaction result is statistical and does not by itself demonstrate biological or pharmacological synergy. In the interaction model, vitamin E was associated with a lower expected top-quality-embryo count when coenzyme Q10 was absent, but the researchers cautioned that this conditional estimate should not be read as evidence that vitamin E generally harms embryo quality. Morphological grading is also an intermediate outcome: it does not fully describe an embryo’s chromosomal status, implantation competence or ultimate potential to result in live birth. The investigators noted that the positive interaction provided a reason to examine the combination in prospective studies, not a basis for inferring a confirmed mechanism in human follicles or embryos.

Clinical outcomes did not mirror the laboratory differences. Pregnancy data were available for 940 of the 949 treatment cycles. Clinical pregnancy occurred in 247 of 940 cycles: 89 of 338 controls, 76 of 260 patients using both supplements, 50 of 224 using coenzyme Q10 alone and 32 of 118 vitamin E-only patients with available data. The corresponding rates were 26.3%, 29.2%, 22.3% and 27.1%, with no statistically significant overall difference. Live birth occurred in 74 of 338 controls, 65 of 260 combined-treatment patients, 42 of 224 coenzyme Q10-only patients and 27 of 118 vitamin E-only patients, giving rates of 21.9%, 25.0%, 18.8% and 22.9%, respectively, with no significant differences. Early-miscarriage rates ranged from 3.6% to 4.4% across the groups and likewise did not differ significantly. Among the 247 clinical pregnancies, 39 ended in early miscarriage, a conditional proportion of 15.8%.

Several limitations constrain the findings. Supplement selection depended on physician judgment and patient acceptance, leaving potential residual confounding from diet, lifestyle, adherence, male-partner factors and other unmeasured characteristics. Nineteen of 968 otherwise eligible cycles were cancelled or interrupted before embryological outcomes could be assessed and were excluded from the main analysis, creating possible selection bias even though cancellation rates did not differ significantly. Baseline vitamin E, coenzyme Q10 and oxidative-stress levels were not measured. Embryo grading may vary between observers, and the records lacked embryonic euploidy, blastocyst-formation and cumulative live-birth data. No single primary embryological outcome was prespecified, and the assessment of multiple outcomes and pairwise comparisons makes the findings exploratory. The authors therefore frame the results as a basis for future multicenter randomized trials rather than a reason for routine supplementation during IVF/ICSI.

Subject of Research: Vitamin E and coenzyme Q10 pretreatment and embryo counts in IVF patients with diminished ovarian reserve

Article Title: Association of vitamin E and coenzyme Q10 pretreatment with embryological and pregnancy outcomes in women with diminished ovarian reserve: a retrospective study

Article References: Wang, S., Wang, T., Zhang, W., Liu, Y., & Cao, Y. (2026). Association of vitamin E and coenzyme Q10 pretreatment with embryological and pregnancy outcomes in women with diminished ovarian reserve: a retrospective study. Frontiers in Endocrinology, 17, Article 1936380. https://doi.org/10.3389/fendo.2026.1936380

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1936380

Keywords: diminished ovarian reserve, in vitro fertilization, intracytoplasmic sperm injection, coenzyme Q10, vitamin E, embryo quality, transferable embryos, pregnancy outcomes, retrospective study, Association, vitamin, coenzyme