Impact of vitamin D levels on ovarian reserve and ovarian response to ovarian stimulation in oocyte donors

Accepted Manuscript Title: Impact of vitamin D levels on ovarian reserve and ovarian response to ovarian stimulation in oocyte donors Author: Alberto Maria Fabris, Maria Cruz, Carlos Iglesias, Alberto Pacheco, Azadeh Patel, Jayesh Patel, Human Fatemi, Juan Antonio García-Velasco PII: DOI: Reference:

S1472-6483(17)30235-3 http://dx.doi.org/doi: 10.1016/j.rbmo.2017.05.009 RBMO 1741

To appear in:

Reproductive BioMedicine Online

Received date: Revised date: Accepted date:

1-9-2016 5-5-2017 9-5-2017

Please cite this article as: Alberto Maria Fabris, Maria Cruz, Carlos Iglesias, Alberto Pacheco, Azadeh Patel, Jayesh Patel, Human Fatemi, Juan Antonio García-Velasco, Impact of vitamin D levels on ovarian reserve and ovarian response to ovarian stimulation in oocyte donors, Reproductive BioMedicine Online (2017), http://dx.doi.org/doi: 10.1016/j.rbmo.2017.05.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Short title: Vitamin D and oocyte donation

Impact of vitamin D levels on ovarian reserve and ovarian response to ovarian stimulation in oocyte donors Alberto Maria Fabris a,*, Maria Cruz a, Carlos Iglesias a, Alberto Pacheco a, Azadeh Patel b , Jayesh Patel b, Human Fatemi c, Juan Antonio García-Velasco a a IVI Madrid, Rey Juan Carlos University, Madrid 28023, Spain; b Department of Reproductive Medicine, Nova IVI Fertility, Ahmedabad, India; c IVI Middle East-Abu Dhabi, UAE * Corresponding author. E-mail address: [email protected] (AM Fabris).

Key message This work supports the assumption that there is not enough evidence to recommend routine vitamin D screening and supplementation prior to assisted reproductive treatment in Caucasian patients. According to current evidence, ovarian reserve markers, ovarian response and cycle outcome of oocyte recipients are unaffected by donors’ vitamin D status.

Page 1 of 13

Dr Fabris graduated as a Medical Doctor and practised as a resident in gynaecology at Padua University, Italy. During 2014 she was Assistant Fellow at Instituto Valenciano de Infertilidad, IVI Madrid, Spain. Now she is working as a gynaecologist at IVI Madrid. She is the author and co-author of several publications.

Abstract Recently, growing interest in vitamin D has emerged from findings that demonstrate a low vitamin D status in populations. Similarly, much interest has been shown in the role that anti-Müllerian hormone (AMH) plays in reproductive physiology. Considerable confusion as to whether vitamin D status is related to ovarian function can be found in the literature. Our retrospective study was performed from June 2014 to April 2015. Oocyte donors were recruited and stimulated under the antagonist protocol with gonadotrophin-releasing hormone (GnRH) agonist to trigger ovulation. In 851 stimulation cycles, we determined the association among serum total and bioavailable vitamin D levels, ovarian reserve and response to ovarian stimulation and the reproductive outcome in their recipients. We showed that vitamin D levels were unrelated to ovarian reserve or ovarian response after ovarian stimulation; in oocyte recipients, gestational outcome did not differ according to a donor’s vitamin D serum status. No correlation was observed between serum AMH and vitamin D. Bioavailable vitamin D was not related to recipients’ ongoing pregnancy rate. Highly prevalent vitamin D insufficiency neither impaired ovarian reserve nor response or oocyte quality in egg donors. No evidence was found for recommending the analysis of vitamin D status in oocyte donors. Keywords: oocyte donation, ovarian reserve, ovarian stimulation, vitamin D

Introduction In the last decade, vitamin D has been in the spotlight in many research fields. Yet despite numerous publications, its influence on reproductive health remains ambiguous. Vitamin D is a steroid hormone, synthesized mainly by the skin upon exposure to ultraviolet light, with small amounts (10–20%) obtained from dietary intake (Wagner et al., 2012). In southern European countries, vitamin D insufficiency is endemic (Hossein-nezhad and Holick, 2013). Recent studies suggest that serum total vitamin D is unable to reflect the real status of this vitamin. In fact, vitamin Dbinding protein values vary and only the bioavailable fraction exerts its function on vitamin D receptors (Powe et al., 2013). For this very reason, bioavailable vitamin D should be considered the best marker of vitamin D status. Anti-Müllerian hormone (AMH) is an established marker of ovarian reserve (Iliodromiti et al., 2014; Peluso et al., 2014). It is strongly associated with oocyte yield after ovarian stimulation and predicts ovarian response to ovarian stimulation (Broer et al., 2011). Although a relationship between vitamin D and AMH has been described (Li, 2014), very little is known about the physiological basis of this correlation. In animal and human models, vitamin D seems to affect AMH gene expression with consequences on granulosa cell differentiation and follicular maturation (Merhi et al., 2014; Irani and Merhi, 2014). Evidence from human IVF is sparse and conflicting. Some authors have described that clinical pregnancy rates progressively lower with declining vitamin D levels (Paffoni et al., 2014; Polyzos et al., 2014; Rudick et al., 2012). On the contrary, other researchers have failed to demonstrate any clear association between vitamin D and

Page 2 of 13

IVF clinical outcome (Aleyasin et al., 2011; Firouzabadi et al., 2014; Franasiak et al., 2015; van de Vijver et al., 2016). Although the effect, if any, could be due to an impact of low vitamin D levels on the oocyte or endometrium, recent evidence seems to suggest that endometrial receptivity is not impaired in women with low vitamin D levels (Fabris et al., 2014). We designed this study to understand the association between total and bioavailable vitamin D (Powe et al., 2013; Yousefzadeh et al., 2014) and ovarian reserve markers: antral follicular count and AMH. Moreover, we analysed ovarian response to ovarian stimulation in relation to vitamin D status. Finally, we compared the gestational outcome of donated egg recipients according to the vitamin D status of their donors. We sought to show whether vitamin D insufficiency or deficiency could be associated with poorer oocyte and embryo quality.

Materials and methods We performed a retrospective study based on the data collected from IVI Madrid over the June 2014 to April 2015 period. All the procedures were approved by an Institutional Review Board (1406-MAD-035-JG) and complied with Spanish law on assisted reproductive technologies (14/2006).

Study population This large non-interventional retrospective study included 851 oocyte donors submitted to 851 ovarian stimulation cycles. Oocyte donors were healthy women aged 18–35 years with regular menstrual cycles, no hereditary or chromosomal diseases, who had normal karyotype and were negative for sexually transmitted diseases (Garrido et al., 2002). Inclusion in the oocyte donor pool also required the donor to have at least six antral follicles per ovary at the beginning of the cycle. Any donors who had polycystic ovary syndrome based on Rotterdam criteria (Azziz, 2006) were excluded. All the donors were submitted to a short antagonist stimulation protocol. An oral contraceptive pill (Microgynon®, levonorgestrel 0.15 mg/ethinylestradiol 0.03 mg, Bayer Hispania, Spain) was taken between 12 and 16 days, starting on day 1 or 2 of the menses of the previous cycle. Following a 5-day wash-out period after the last pill, donors started the stimulation protocol. They received daily doses of 150/225 IU of recombinant FSH (Gonal-f®; Merck Serono, Spain) depending on body mass index (BMI) and antral follicle count. Daily doses of 0.25 mg gonadotrophin-releasing hormone (GnRH) antagonist (Cetrotide®; Merck Serono, Spain) were started when one follicle reached a mean diameter of 13 mm. A single dose of 0.2 mg GnRH agonist (Decapeptyl®; Ipsen Pharma, Spain) was administered to trigger final oocyte maturation when at least three follicles reached a mean diameter of 17 mm. Oocyte pick-up was performed under sedation 36 h later. Oocyte recipients ranged from 29 to 49 years old with a BMI of 18–38 kg/m2. Exclusion criteria were endometriosis stage III–IV, hydrosalpinx, acquired or congenital uterine abnormalities, and partner’s azoospermia. According to the endometrial preparation for oocyte recipients, the protocol was as follows: women with ovarian function were down-regulated in the luteal phase with a single dose of GnRH agonist depot (Decapeptyl® 3.75 mg; Ipsen Pharma, Spain or Gonapeptyl® 3.75 mg; Ferring, Spain). After menses, all the subjects received oral oestradiol valerate (Progynova®; Schering-Plough, Spain or Meriestra®; Novartis Farmaceutica, Spain), which started at a daily dose of 2 mg and was increased gradually to 6 mg.

Page 3 of 13

Approximately 10–15 days after initiating oestradiol valerate, serum oestradiol levels and endometrial thickness were determined. Administration of micronized progesterone (800 mg/day, vaginally) was initiated the day after oocyte donation (Progeffik®; Effik Laboratories, Spain or Utrogestan®; SEID, Spain). Donors’ oocytes were fertilized by intracytoplasmic sperm injection (ICSI) according to our internal protocol in the egg donation programme. In all cases, fresh embryo transfer was performed and excess embryos were vitrified. One or two top-quality embryos were transferred according to selection by gynaecologists and patients. The morphological embryonic characteristics that were taken into account were the degree of blastocyst expansion quality of inner cell mass and trophectoderm according to the Gardner scoring system (Gardner et al., 2000). Each recipient received eggs from one donor only and each donor donated to one recipient. Eggs of donors without compatible recipients were vitrified. Of the 679 compatible recipients, no fresh embryo transfer was performed in 12 cases due to insufficient endometrial thickness or vaginal bleeding of the recipient. Therefore, out of 851 donors the gestational outcome of their recipients and correlation of the outcome with the donors’ vitamin D status was analysed in 667 cases. Implantation rate was calculated as the number of intrauterine gestational sacs observed by transvaginal ultrasonography divided by the number of transferred embryos. Clinical pregnancy was defined as a pregnancy with ultrasound visualization of one or more gestational sacs with foetal heartbeat. Clinical pregnancy rate was calculated as the number of clinical pregnancies per number of embryo transfer cycles. Miscarriage rate is the number of pregnancies lost before 24 weeks of gestation divided by the number of clinical pregnancies. Finally, ongoing pregnancy rate is defined as a pregnancy that has progressed beyond week 12 divided by the number of transfers performed.

Hormone assays Blood samples were allowed to clot, and sera were separated by centrifugation and stored at –80C until analyses were performed. Serum AMH levels were determined by a single measurement with the Elecsys AMH assay in an e411 Cobas automated device (Roche). The limit of functional sensitivity was 0.03 ng/ml and the coefficient of variation was 3.5%, as determined for the control samples during the study. Oestradiol levels on the day of ovulation induction were determined by chemiluminescence in an automatic analyser Architect i100 (Abbot Diagnostics). The intra- and inter-assay coefficients of variation were below 6% in both cases. For oestradiol, the limit of detection (LoD) was <10 pg/ml. A Multiscan ELISA reader (Labsystems) measured vitamin D-binding protein (VDBP) using a commercial ELISA kit (R&D Systems) following the manufacturer’s recommendations. The LoD for VDBP was 0.15 ng/ml, the linearity limit was 250 ng/ml, and the intra-/inter-assay coefficients of variation were 5.8% and 6.0%, respectively. Vitamin D was determined by chemiluminescence in an Advia Centaur device (Siemens). The LoD for vitamin D was 3.2 ng/ml, the linearity limit was 150 ng/ml, and the intra-/inter-assay coefficients of variation were 7.0% and 11.1%, respectively. Albumin was measured with a Cobas Mira Plus analyser (Roche). The LoD for albumin was 0.04 g/dl, the linearity limit was 6.0 g/dl, and the intra-/interassay coefficients of variation were 0.52% and 0.78%, respectively. Finally, bioavailable vitamin D was calculated using Vermeulen’s validated formula (Vermeulen, 1999).

Page 4 of 13

Vitamin D repletion, insufficiency and deficiency were defined according to international consensus. The Endocrine Society has defined vitamin D insufficiency as serum levels between 20 and 30 ng/ml, and vitamin D deficiency as levels below 20 ng/ml (Holick et al., 2011), while bioavailable vitamin D was considered a continuous variable.

Statistical analysis The data from the clinical outcomes are presented as descriptive statistics (mean ± SEM). The clinical results were analysed by a t-test to compare the means and chisquared tests for proportions. Receiver operating characteristic (ROC) curves were used to analyse the predictive value of bioavailable vitamin D on the ongoing pregnancy rate. The database from this retrospective study included 851 cycles from 851 donors. With this sample size, the statistical power to detect a 20% improvement in a unilateral test was 85.3%, by considering a significance level of 95%. In order to look for any dependence or statistical association between the bioavailable vitamin D levels and the ovarian reserve/response markers, we did a correlation analysis to help us to indicate a predictive relationship to be exploited in practice. A P-value <0.05 was considered statistically significant. Statistical analyses were performed using the Statistical Package for Social Sciences 19.0 (IBM Corporation, New York, USA).

Results This retrospective study included 851 oocyte donors. Of the 851 controlled ovarian stimulation cycles, 29.5% of donors (n = 251) were vitamin D replete (vitamin D ≥30 ng/ml), 52% (n = 443) had vitamin D insufficiency (vitamin D 20–30 ng/ml) and 18.5% (n = 157) exhibited vitamin D deficiency (<20 ng/ml).

Oocyte donors and recipients’ outcomes The clinical characteristics of the oocyte donor population are presented in Table 1. Age and BMI were statistically and significantly different, but no clinically relevant differences among the three study groups were seen. According to ovarian reserve stratifying on vitamin D levels, serum AMH and antral follicle count (AFC) were similar among the donors with normal, insufficient or deficient total vitamin D. Similarly, the variables that related to ovarian response, e.g. oestradiol levels on the day of GnRH agonist administration, number of oocytes retrieved and mature oocytes, were comparable. Regarding recipients’ characteristics, the baseline parameters are shown in Table 2. At this point, the number of transferred embryos slightly differed and was statistically significantly different (P = 0.027), but the difference was not clinically relevant. The total frozen embryos per cycle did not differ in relation to the donor’s vitamin D class. The clinical outcomes for the three groups were also similar (Table 3).

Serum bioavailable vitamin D and markers of ovarian reserve (AMH and AFC) In order to measure whether there was any relationship between two random quantitative variables, we applied the Pearson correlation coefficient (r). The obtained data were as follows: the value between the AMH levels and the bioavailable vitamin D levels was r = 0.059. The relationship between AFC and vitamin D in its bioavailable form yielded a value of r = 0.081 (P = 0.011). Regarding ovarian response, the results mirrored previous data; for the retrieved oocytes and bioavailable vitamin D, Page 5 of 13

r = –0.002; finally, the Pearson correlation coefficient between metaphase II oocytes and bioavailable vitamin D was r = –0.053. In none of the assessed cases did the Pvalue show a significant correlation between the two analysed parameters. We also performed a regression model to adjust for potential confounders as BMI index and age, and we found no significant differences among the three study groups. The predictive values of both total vitamin D (Figure 1) and bioavailable vitamin D (Figure 2) regarding ongoing pregnancy rates in the 667 recipients who underwent embryo transfer cycles with fresh embryos were analysed by a ROC curve. For the former, the area under the curve (AUC) was 0.503 (95% confidence interval (CI), 0.457–0.543), and was AUC = 0.554 (95% CI, 0.490–0.617) for the latter, which indicates that the AUC analysis for both parameters was not informative.

Discussion The issue as to whether vitamin D levels are associated with gestational outcomes in assisted reproductive outcomes is still a controversial matter. Evidence is still poor as no randomized controlled trials are currently available, and the results of existing small cohorts are highly heterogeneous. Given all these problems, it is difficult to accurately advise vitamin D supplementation to vitamin D-depleted women who undergo assisted reproductive treatment. In our study, we failed to find any positive correlation between low serum vitamin D levels and poorer IVF outcome. Our results are in line with those of other groups, who were also unable to find any relationship between vitamin D levels and clinical outcomes in women who underwent assisted reproductive treatment (Firouzabadi et al., 2014; Franasiak et al., 2015; van de Vijver et al., 2016). On the contrary, our results contrast with other literature reports that have found a strong negative effect of poor vitamin D levels on assisted reproductive technology outcome (Ozkan et al., 2010; Rudick et al., 2012, 2014). Polyzos et al. (2014) reported a higher rate of elective single-embryo transfer (SET) in patients with normal vitamin D levels, maybe due to a higher embryo quality among which they could choose, so they assumed that vitamin D could play a role in folliculogenesis, oogenesis and embryo quality. In order to find reasons to help explain these discrepancies with previous studies, we looked for any possible methodological differences among them. First, most previous studies have included a much smaller population, so their results are exposed to a type II error. Our sample was significantly larger, which offers greater reassurance about the validity of our findings. Second, ethnic and regional differences may partially explain different findings. We reduced this confounder as our patients were 95% Caucasian. Early retrospective studies have postulated that vitamin D deficiency might negatively affect pregnancy rates with an effect mediated through the endometrium (Rudick et al., 2014). However, we were unable to confirm this hypothesis using the oocyte– donor recipient model (Fabris et al., 2014). Another recent report (van de Vijver et al., 2016) has concluded that vitamin D deficiency does not significantly impair pregnancy rates in those patients who have undergone frozen-thawed embryo transfer. Thus, endometrial receptivity does not seem to be affected by vitamin D status.

Page 6 of 13

The main strength of our study is that it was designed to evaluate not only vitamin D, but also bioavailable vitamin D levels, a biomarker that is more robust than total vitamin D (Powe et al., 2013). The free hormone hypothesis states that protein-bound hormones are relatively inactive, while hormones not bound to binding proteins are available to exert biological activity (Mendel, 1989). The majority of circulating vitamin D is tightly bound to VDBP with less than 1% of vitamin D circulating as a free form and exerting its action (Chun et al., 2014). Finally, it is necessary to highlight that the major advantage of providing bioavailable vitamin D levels is that it can allow us to control for differences in VDBP between different races, and even if just 5% of our population is not Caucasian. As previously outlined, the existing literature provides no consistent pattern for how vitamin D may influence AMH secretion (Dennis et al., 2012; Merhi et al., 2012). If vitamin D increases AMH production, theoretically it is possible that vitamin D supplementation could help delay ovarian reserve reduction by producing significant reproductive and general health benefits (Schoenaker et al., 2014). Hence one of the key findings in our large retrospective study is that serum vitamin D levels appear to be unrelated to AMH values, as we observed similar concentrations of this hormone regardless of the vitamin D category. What is more, these data are fortified because no significant differences were found in either AFC or the number of oocytes retrieved during follicular retrieval. Although these results stem from a very homogenous study population in which baseline characteristics were controlled, they agree with a previous study (Pearce et al., 2015). These authors concluded that serum AMH levels were not associated with serum vitamin D concentrations, not even after controlling for relevant covariants. The strengths of the present study include, according to our knowledge, the largest dataset from a single centre to have analysed the influence of vitamin D on ovarian reserve and response using an oocyte donor–recipient model. We wish to emphasize the importance of these results in this context as interest in vitamin D has grown in the past few years in lay and biomedical literature. Despite the advantages that our dataset confers the analysis, limitations still remain. First, as a consequence of being a retrospective study, not all the pertinent risk factors are likely to have been identified and subsequently recorded. So only association, and not causation, can be inferred from the results. Second, a further limitation of the study is that donors are healthy young women and do not represent other populations, especially infertile patients who are often low responders and older, or present polycystic ovary syndrome or other pathologies. However, this population allows the exclusion of confounding factors and concomitant diseases that could bias the results. Egg donors were present in large percentages of insufficient or deficient values of serum vitamin D, as already described in general southern European populations and in our infertile patients, recipients of egg donations. Whatever the vitamin D status, no influence was shown in ovarian reserve and ovarian response to stimulation. Moreover, egg donors presented a wide range of AMH values and of number of retrieved oocytes. If vitamin D status was strongly correlated to either AMH or number of oocytes obtained, it would be possible to find a correlation even in this population, which was not the case. Another limitation is that the vast majority of patients were Caucasian and the results cannot be extended to other racial groups. Moreover, the fact that we measured bioavailable vitamin D weakened the effect of VDBP polymorphisms that could be due to racial differences (Powe et al., 2013). In short, the results from this work support the assumption that there is not enough strong evidence to support recommending routine vitamin D screening and

Page 7 of 13

supplementation prior to assisted reproductive treatment in Caucasian patients. Perhaps extremely low vitamin D levels could have a stronger influence on reproduction but, according to current evidence, ovarian reserve markers, ovarian response and cycle outcome of oocyte recipients are unaffected by donors’ vitamin D status. There is a need for trials in this field to investigate any therapeutic benefits of vitamin D supplementation.

Acknowledgements The authors thank Monica Toribio for her support in managing this paper. This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References Aleyasin, A., Hosseini, M.A., Mahdavi, A., Safdarian, L., Fallahi, P., Mohajeri, M.R., Abbasi, M. and Esfahani, F. 2011. Predictive value of the level of vitamin D in follicular fluid on the outcome of assisted reproductive technology. European Journal of Obstetrics, Gynecology, and Reproductive Biology 159, 132–137. Azziz, R. 2006. Controversy in clinical endocrinology: diagnosis of polycystic ovarian syndrome: the Rotterdam criteria are premature. The Journal of Clinical Endocrinology and Metabolism 91, 781–785. Broer, S.L., Dolleman, M., Opmeer, B.C., Fauser, B.C., Mol, B.W. and Broekmans, F.J. 2011. AMH and AFC as predictors of excessive response in controlled ovarian hyperstimulation: a meta-analysis. Human Reproduction Update 17, 46–54. Chun, R.F., Peercy, B.E., Orwoll, E.S., Nielson, C.M., Adams, J.S. and Hewison, M. 2014. Vitamin D and DBP: the free hormone hypothesis revisited. The Journal of Steroid Biochemistry and Molecular Biology 144, 132–137. Dennis, N.A., Houghton, L.A., Jones, G.T., van Rij, A.M., Morgan, K. and McLennan, I.S. 2012. The level of serum anti-Mullerian hormone correlates with vitamin D status in men and women but not in boys. The Journal of Clinical Endocrinology and Metabolism 97, 2450–2455. Fabris, A., Pacheco, A., Cruz, M., Puente, J.M., Fatemi, H. and Garcia-Velasco, J.A. 2014. Impact of circulating levels of total and bioavailable serum vitamin D on pregnancy rate in egg donation recipients. Fertility and Sterility 102, 1608–1612. Firouzabadi, R.D., Rahmani, E., Rahsepar, M. and Firouzabadi, M.M. 2014. Value of follicular fluid vitamin D in predicting the pregnancy rate in an IVF program. Archives of Gynecology and Obstetrics, 289, 201–206. Franasiak, J.M., Molinaro, T.A., Dubell, E.K., Scott, K.L., Ruiz, A.R., Forman, E.J., Werner, M.D., Hong, K.H. and Scott, R.T.,Jr 2015. Vitamin D levels do not affect IVF outcomes following the transfer of euploid blastocysts. American Journal of Obstetrics and Gynecology 212, 315.e1–315.e6. Gardner DK, Lane M, Stevens J, Schoolcraft WB. 2000. Blastocyst score affects implantation and pregnancy outcome: towards a single blastocyst transfer. Fertility and Sterility 73, 1155–1158. Garrido, N., Zuzuarregui, J.L., Meseguer, M., Simon, C., Remohi, J. and Pellicer, A. 2002. Sperm and oocyte donor selection and management: experience of a 10-year follow-up of more than 2100 candidates. Human Reproduction (Oxford, England) 17, 3142–3148. Hossein-nezhad, A. and Holick, M.F. 2013. Vitamin D for health: a global perspective. Mayo Clinic Proceedings 88, 720–755. Iliodromiti, S., Kelsey, T.W., Wu, O., Anderson, R.A. and Nelson, S.M. 2014. The predictive accuracy of anti-Mullerian hormone for live birth after assisted conception: a systematic review and metaanalysis of the literature. Human Reproduction Update 20, 560–570. Irani, M. and Merhi, Z. 2014. Role of vitamin D in ovarian physiology and its implication in reproduction: a systematic review. Fertility and Sterility 102, 460–468.e3. Mendel, C.M. 1989. The free hormone hypothesis: a physiologically based mathematical model. Endocrine Reviews 10, 232–274. Merhi, Z., Doswell, A., Krebs, K. and Cipolla, M. 2014. Vitamin D alters genes involved in follicular development and steroidogenesis in human cumulus granulosa cells. The Journal of Clinical Endocrinology and Metabolism 99, E1137–45. Merhi, Z.O., Seifer, D.B., Weedon, J., Adeyemi, O., Holman, S., Anastos, K., Golub, E.T., Young, M., Karim, R., Greenblatt, R. and Minkoff, H. 2012. Circulating vitamin D correlates with serum antimullerian hormone levels in late-reproductive-aged women: Women’s Interagency HIV Study. Fertility and Sterility 98, 228–234.

Page 8 of 13

Ozkan, S., Jindal, S., Greenseid, K., Shu, J., Zeitlian, G., Hickmon, C. and Pal, L. 2010,. Replete vitamin D stores predict reproductive success following in vitro fertilization. Fertility and Sterility 94, 1314–1319. Paffoni, A., Ferrari, S., Vigano, P., Pagliardini, L., Papaleo, E., Candiani, M., Tirelli, A., Fedele, L. and Somigliana, E. 2014. Vitamin D deficiency and infertility: insights from in vitro fertilization cycles. The Journal of Clinical Endocrinology and Metabolism 99, E2372–6. Pearce, K., Gleeson, K. and Tremellen, K. 2015. Serum anti-Mullerian hormone production is not correlated with seasonal fluctuations of vitamin D status in ovulatory or PCOS women. Human Reproduction (Oxford, England) 30, 2171–2177. Peluso, C., Fonseca, F.L., Rodart, I.F., Cavalcanti, V., Gastaldo, G., Christofolini, D.M., Barbosa, C.P. and Bianco, B. 2014. AMH: An ovarian reserve biomarker in assisted reproduction. Clinica Chimica Acta; International Journal of Clinical Chemistry 437, 175–182. Polyzos, N.P., Anckaert, E., Guzman, L., Schiettecatte, J., Van Landuyt, L., Camus, M., Smitz, J. and Tournaye, H. 2014. Vitamin D deficiency and pregnancy rates in women undergoing single embryo, blastocyst stage, transfer (SET) for IVF/ICSI. Human Reproduction (Oxford, England) 29, 2032–2040. Powe, C.E., Evans, M.K., Wenger, J., Zonderman, A.B., Berg, A.H., Nalls, M., Tamez, H., Zhang, D., Bhan, I., Karumanchi, S.A., Powe, N.R. and Thadhani, R. 2013. Vitamin D-binding protein and vitamin D status of black Americans and white Americans. The New England Journal of Medicine 369, 991–2000. Rudick, B., Ingles, S., Chung, K., Stanczyk, F., Paulson, R. and Bendikson, K. 2012. Characterizing the influence of vitamin D levels on IVF outcomes. Human Reproduction (Oxford, England) 27, 3321–3327. Rudick, B.J., Ingles, S.A., Chung, K., Stanczyk, F.Z., Paulson, R.J. and Bendikson, K.A. 2014. Influence of vitamin D levels on in vitro fertilization outcomes in donor-recipient cycles. Fertility and Sterility 101, 447–452. Schoenaker, D.A., Jackson, C.A., Rowlands, J.V. and Mishra, G.D. 2014. Socioeconomic position, lifestyle factors and age at natural menopause: a systematic review and meta-analyses of studies across six continents. International Journal of Epidemiology 43, 1542–1562. van de Vijver, A., Drakopoulos, P., Van Landuyt, L., Vaiarelli, A., Blockeel, C., Santos-Ribeiro, S., Tournaye, H. and Polyzos, N.P. 2016. Vitamin D deficiency and pregnancy rates following frozen-thawed embryo transfer: a prospective cohort study. Human Reproduction (Oxford, England) 31, 1749–1754. Wagner, C.L., Taylor, S.N., Dawodu, A., Johnson, D.D. and Hollis, B.W. 2012. Vitamin D and its role during pregnancy in attaining optimal health of mother and fetus. Nutrients 4, 208–230. Yousefzadeh, P., Shapses, S.A. and Wang, X. 2014. Vitamin D Binding Protein Impact on 25Hydroxyvitamin D Levels under Different Physiologic and Pathologic Conditions. International Journal of Endocrinology 2014,981581.

Declaration: The authors report no financial or commercial conflicts of interest.

Figure 1. Receiver operating characteristic (ROC) curve for assessment of the predictive value of total vitamin D for ongoing pregnancy rate; AUC = 0.503 (95% CI, 0.457–0.543). Figure 2. Receiver operating characteristic (ROC) curve for assessment of the predictive value of bioavailable vitamin D for ongoing pregnancy rate; AUC = 0.554 (95% CI, 0.490–0.617).

Page 9 of 13

Table 1 – Clinical characteristics of the oocyte donors. Characteristics

Vitamin D replete (≥30 ng/ml)

Vitamin D insufficiency (20–30 ng/ml)

Vitamin D deficiency (≤20 ng/ml)

n

251

443

157

Agea

25.3 ± 0.2

24.7 ± 0.3

25.0 ± 0.4

BMI (kg/m2)b

21.9 ± 0.2

22.9 ± 0.2

23.1 ± 0.5

10.5 ± 0.2

10.6 ± 0.1

10.3 ± 0.2

2.4 ± 0.2

2.3 ± 0.1

2.3 ± 0.3

Antral follicles

22.0 ± 0.6

22.0 ± 0.5

21.7 ± 1.3

Oestradiol (pg/ml)

3578 ± 567

3387 ± 321

3479 ± 487

Oocytes retrieved

17.5 ± 1.0

17.1 ± 0.6

16.5 ± 1.0

MII oocytes

12.4 ± 0.9

12.5 ± 0.2

13.0 ± 0.1

Days of stimulation AMH (ng/ml) c

AMH = anti-Müllerian hormone; BMI = body mass index; MII = metaphase II. a P = 0.034. b P < 0.001. c On day of GnRH agonist trigger administration.

Page 10 of 13

Table 2 – Characteristics of the oocyte recipients. Vitamin D replete donors (≥30 ng/ml)

Vitamin D insufficient donors (20–30 ng/ml)

Vitamin D deficient donors (≤20 ng/ml)

n

193

379

107

Age

40.6 ± 0.5

41.0 ± 0.4

40.7 ± 0.7

BMI (kg/m2) Transferred embryosa Frozen embryos

22.7 ± 0.5

22.7 ± 0.5

22.8 ± 0.8

1.6 ± 0.05

1.7 ± 0.05

1.6 ± 0.1

3.5 ± 0.3

3.7 ± 0.3

3.6 ± 0.4

Characteristics

BMI = body mass index. a P = 0.027.

Page 11 of 13

Table 3 – Clinical outcomes of the oocyte donation cycles.

Receptors

Vitamin D replete donors (≥30 ng/ml)

Vitamin D insufficient donors (20–30 ng/ml)

Vitamin D deficient donors (≤ 20 ng/ml)

Implantation rate (%)

46.7 (148/317)

46.6 (304/653)

52.4 (89/170)

Clinical pregnancy rate (%)

66.3 (128/193)

64.3 (240/373)

75.2 (76/101)

Miscarriage rate (%)

23.4 (30/128)

25.0 (60/240)

28.9 (22/76)

Ongoing pregnancy rate (%)

50.8 (98/193)

48.3 (180/373)

53.5 (54/101)

There were no statistically significant differences between the groups.

Page 12 of 13

Figure 1.

Figure 2.

Page 13 of 13