{"paper_id":"3add3d0b-e38c-45e5-94c9-9b20c037a76d","body_text":"Vitamin D deficiency is a ‘global pandemic’ that affects all populations including reproductive-age women [ 1 ,  2 ]. While vitamin D is well known to be important for bone health, vitamin D receptors and vitamin D metabolizing enzymes have also been found in various reproductive organs including the ovaries, uterus, endometrium, and placenta and thus postulated to play a role in normal reproductive function [ 3 ].\nBasic science studies have shown that vitamin D can affect ovarian follicular development and luteinization by altering AMH signalling, FSH sensitivity, and progesterone production and release in the human granulosa cells [ 4 ]. Other studies have suggested a possible effect of vitamin D on endometrial receptivity via regulating myometrial contraction and the expression of the homeobox gene HOXA10, which is a well-known molecule involved in the mechanism of implantation [ 5 ,  6 ]. In addition, vitamin D receptors have been found in maternal decidua and foetal trophoblast in early pregnancy, suggesting a role in trophoblast invasion, placental spiral artery remodelling, and regulating immune cell function [ 7 ,  8 ]. There is still controversy on how these mechanisms translate exactly to affect clinical reproductive outcome.\nIn vitro fertilization (IVF) provides a good model to study the individual steps in human reproduction. In recent years, meta-analyses have generally shown that vitamin D deficiency in women undergoing IVF was associated with a lower live birth rate compared to women who were vitamin D replete, although data is still not consistent [ 9 – 12 ]. Many of the studies assessed the vitamin D levels and pregnancy outcome at the fresh IVF cycle. High oestradiol levels during ovarian stimulation for IVF can stimulate increase in vitamin D-binding protein and affect the total vitamin D serum concentrations [ 13 ]. Moreover, embryo cryopreservation has become an integral part of modern-day IVF and many women may have all embryos frozen without fresh embryo transfer due to different clinical indications [ 14 ]. A retrospective study performed at our unit showed that the cumulative live birth rate of the first IVF cycle in the vitamin D deficient group was significantly lower compared to the non-deficient group [ 15 ]. However, serum vitamin D levels were assessed at the start of ovarian stimulation for the IVF cycle in our study and frozen embryo transfer could have occurred months later in some individuals.\nStudies of IVF using donor oocytes enable factors that impact implantation (recipient factors) to be analysed separately from factors that influence ovarian stimulation and embryo quality (donor factors). Women who donate oocytes are presumably younger and have good oocyte quality, therefore the studied effect on recipients may be more related to endometrial factors. One retrospective cohort study involving 99 recipients of fresh embryo transfer from donor oocytes showed improvement in pregnancy rate in recipient women who were vitamin D replete compared to those who were deficient (live birth rates 59% vs 31%) and supported that the effect of vitamin D may be mediated via enhancing endometrial receptivity [ 16 ]. However, three other studies did not show any differences between pregnancy outcome in vitamin D deficient and replete recipients of embryos from donor oocytes [ 17 – 19 ] so the effect of vitamin D on endometrial receptivity is still controversial.\nOnly a few reports have studied the association of vitamin D and pregnancy outcomes in frozen embryo transfer and the results are again conflicting. A retrospective study of 368 infertile women undergoing single frozen blastocyst transfer showed that vitamin D deficiency compromised pregnancy rates by 40% [ 20 ]. Franasiak et al. further attempted to control for embryonic factors in another retrospective cohort study by including only euploid blastocyst transfers and did not find the serum vitamin D status to be related to pregnancy outcome, although the population was a mix of 517 first fresh or cryopreserved embryo transfers [ 21 ]. A prospective observational cohort study involving 280 patients undergoing frozen embryo transfer also did not show that vitamin D deficiency affected pregnancy rates [ 22 ]. These studies involved mostly Caucasian population and serum vitamin D was assessed using immunoassay.\nIn this study, we assessed the association between serum vitamin D level and the live birth rate in women undergoing frozen embryo transfer.\n\nThis was a retrospective study carried out at 2 tertiary reproductive medicine centres in Hong Kong, namely the Centre of Assisted Reproduction and Embryology, The University of Hong Kong—Queen Mary Hospital and the Dr. Stephen Chow Chun—Kay Assisted Reproduction Centre, Kwong Wah Hospital. Clinical details and laboratory data were prospectively entered into a computerized database by designated embryologists and retrieved for analysis. To avoid transcription error, the data was routinely double-checked by a second designated embryologist. Subjects are identified by three identifiers (name, identification number, and date of birth), and checked against the treatment date. In case of missing data, the centralized Hospital Authority electronic patient record system that links up all public hospitals in Hong Kong and paper records were reviewed to obtain the missing data. The study was approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster and the Kowloon Central Cluster/Kowloon East Cluster Research Ethics Committee.\nWomen included in this retrospective analysis were those who were undergoing frozen embryo transfer at the two IVF centres between 2019 and 2021. Those undergoing in vitro maturation, pre-implantation genetic testing, and whose archived serum sample could not be retrieved were excluded. Those who had repeated IVF or frozen embryo transfer cycles during the period were counted only for the earliest cycle and therefore the first frozen embryo transfer of the first IVF cycle in the study period was used for analysis where possible.\nThe details of the procedures for ovarian stimulation, oocyte retrieval, handling of the gametes, cryopreservation of the embryos, and frozen embryo transfer were previously described [ 15 ,  23 ].\nThe gonadotrophin-releasing hormone (GnRH) antagonist, long GnRH agonist, or progestin-primed protocols were used for pituitary downregulation. Human menopausal gonadotrophin (HMG) or recombinant FSH was used for ovarian stimulation. In the GnRH antagonist protocol, women received ganirelix (Orgalutran®, NV Organon, The Netherlands) or cetrorelix (Cetrotide®, Merck, Germany) 250 µm daily starting from the sixth day of stimulation. In the long GnRH agonist protocol, the women received buserelin (Suprecur®, Hoechst, Frankfurt, Germany) nasal spray 150 µm four times a day starting from the mid-luteal phase of the cycle preceding the treatment cycle. In the progestin-primed ovarian stimulation protocol, 10 mg daily medroxyprogesterone acetate was given simultaneously with gonadotrophins at the start of stimulation [ 24 ]. The initial dose of stimulation was determined according to the baseline antral follicle count (AFC). Human chorionic gonadotrophin (hCG) (Ovidrel® 250 µm) was given when the mean diameter of the leading follicle reached 18 mm and more than 3 follicles reaching a mean diameter of 16 mm or above, followed by transvaginal ultrasound-guided oocyte retrieval 36 h later. For women with excessive ovarian response using the GnRH antagonist or progestin-primed ovarian stimulation protocols as shown by serum oestradiol level of > 25,000 pmol/L, more than 15 follicles larger than 14 mm in diameter or evidence of ovarian hyperstimulation syndrome, GnRH agonist (Decapeptyl®, Ferring, Kiel, Germany) 0.3 mg can be given subcutaneously to replace hCG to trigger final oocyte maturation. Fertilization was carried out either by conventional insemination or intracytoplasmic sperm injection (ICSI) depending on semen parameters.\nPrior to June 2020, cryopreservation of day 2 cleavage stage embryos was performed by a slow freezing protocol using a programmable freezer (Planer Products Ltd.; Sunbury-On-Thames, UK) in Queen Mary Hospital. Vitrification was performed for cleavage-stage embryos in Kwong Wah Hospital and starting from June 2020 in Queen Mary Hospital as well. Vitrification was performed for blastocysts in both hospitals. Extended culture to blastocyst was advised if there were more than 3 good quality cleavage stage embryos 2 days after retrieval.\nFrozen embryo transfer was performed in natural, letrozole-induced or hormone replacement cycles. Frozen embryos or blastocysts were transferred in natural cycles in ovulatory women and in letrozole-induced or hormone-replacement cycles for anovulatory women. Hormone replacement cycles can also be used in ovulatory women who had difficulty in identifying the day of LH surge in previous cycles or for scheduling issues.\nIn natural cycles, women had a daily blood test to identify the day of LH surge starting from 18 days before the next expected period (17). The LH surge was defined by the LH level being above 20 IU/L and more than double the average over the past 3 days. Luteal phase support was not routinely given.\nIn letrozole-induced cycles, women received letrozole 2.5 mg daily for 5 days, followed by daily monitoring of serum LH, similar to that in natural cycles. Letrozole was stepped up to 5–7.5 mg if needed. Luteal phase support was not routinely given.\nIn hormone replacement cycles, women received oral oestradiol 6 mg daily for 14 days for endometrial priming, followed by transvaginal ultrasonography for endometrial thickness and addition of vaginal micronized progesterone 100 mg and oral dydrogesterone 10 mg three times daily if endometrial thickness reached 7 mm or above. If they were pregnant, these women continued oral oestradiol and vaginal/oral progestogen after frozen embryo transfer until 12 weeks of gestation.\nFor natural cycles and letrozole-induced cycles, frozen day 2 or day 3 embryos were transferred on the third or fourth day respectively after the LH surge. For hormone replacement cycles, frozen day 2 or day 3 embryos were transferred on the fourth or fifth day after starting progestogen. Frozen blastocysts were transferred on the sixth day after the LH surge or on the seventh day after starting progestogen.\nWomen were allowed to have a replacement of at most two cleavage stage embryos if she was > 38 years old or had failed 2 previous IVF cycles, and had no live birth, but single embryo transfer was strongly encouraged otherwise. Only single blastocyst transfer was allowed.\nThe embryo transfer was performed under transabdominal ultrasound guidance with a soft catheter (Sydney IVF Embryo Transfer Catheter®, Cook, Indiana, USA).\nUrine pregnancy test was performed 18 days after the LH surge in FET cycles. If the urine pregnancy test was positive, transvaginal scanning was performed at 6 and 8 weeks of gestation to confirm foetal viability and ongoing pregnancy. Pregnancy outcomes were tracked from the Hospital Authority electronic patient record system or self-returned reply slips from the women or their obstetricians. If the woman did not deliver within the public hospital system and no reply letter was received 2–3 months after the expected date of delivery, they were contacted by our nurses to update the database.\nArchived serum samples used for analysis in this study were taken at luteinizing hormone (LH) surge in natural or letrozole-induced cycles or 14 days after the start of oestradiol and at the commencement of progestogens in hormone replacement cycles. Residual serum after the routine clinical tests was archived at − 20 °C with written informed consent from the patients for research purposes anonymously. The average time was around 4 h between drawing the sample and archiving. The archived serum samples were retrieved and assayed for serum 25(OH)D concentration using liquid chromatography-mass spectrometry (LC–MS) at Adicon Clinical Laboratories Limited, which is accredited by the Chinese National Center for Clinical Laboratories (NCCL), External Quality Assessment (EQA), and Trueness Verification of 25(OH)D Assays scheme. The limit of quantification is 2.5 nmol/l (1 ng/ml). The within-run and between-run trueness for low, medium, and high levels of vitamin D2 are 92.0–113.2%, 94.3–106.7%, and 100.4–106.8%, respectively, whereas that for vitamin D3 are 104.7–118.5%, 97.8–109.6%, and 103.7–110.3% respectively. The co-efficient of variation within and between-runs for low, medium, and high levels of vitamin D2 are 4.38–6.05%, 2.24–4.17%, and 1.63–2.03%, respectively, and for vitamin D3 are 0.86–6.33%, 1.12–3.55%, and 0.72–2.85%.\nVitamin D deficiency was defined as serum 25(OH)D levels < 50 nmol/L (< 20 ng/ml) and vitamin D insufficiency as serum 25(OH)D ≥ 50 and < 75 nmol/L (≥ 20 and < 30 ng/ml) in accordance with the Endocrine Society criteria [ 25 ]. Serum 25(OH)D levels of ≥ 75 nmol/L (30 ng/ml) were considered replete. Women who were vitamin D insufficient and replete were grouped together as the non-deficient group and compared with those who were vitamin D deficient in the primary analysis. The three groups (vitamin D deficient, insufficient, and replete) were separately compared in further analysis.\nThe primary outcome was the live birth rate. Live birth was defined as the delivery of an infant born alive after 24 weeks of gestation (the definition of foetal viability adopted in this locality). Other outcomes included the pregnancy rate, ongoing pregnancy rate, multiple pregnancy rate, miscarriage rate, and ectopic pregnancy rate. Ongoing pregnancy was defined as a viable pregnancy beyond 8 weeks of gestation. Miscarriage was defined as the spontaneous loss of clinical pregnancy before 20 completed weeks of gestational age. Ectopic pregnancy was defined as a pregnancy outside the uterine cavity, diagnosed by ultrasound, surgical visualization, or histopathology.\nData were entered and analysed using IBM SPSS 26.0 (IBM Corporation, NY, USA). Continuous and categorical variables were compared between groups using the Mann–Whitney  U  test, Kruskal–Wallis test, and Chi-square test respectively. Odds ratio was calculated with the vitamin D non-deficient group as the reference group. Logistic regression analysis was performed to assess the effect of vitamin D on pregnancy outcomes controlling for confounders including women’s age at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type, and duration of infertility. The two-tailed  P  value of < 0.05 was considered statistically significant.\n\nOne thousand four hundred eighty-nine women who had FET during the study period were available for analysis. The flow of participants is shown in Fig.  1 . For women who had repeat frozen embryo transfer cycles in the study period, only the first frozen embryo transfer was counted and the other duplicate cycles were excluded. Fig. 1 Flowchart of participants\nFlowchart of participants\nThe median age of the women at oocyte retrieval was 36 (25th–75th percentile 34–38) years. A total of 561/1489 (37.7%) women had cleavage stage embryo transfer and 928/1489 (62.3%) women had blastocyst transfer. A total of 1482/1489 (99.5%) women had single cleavage stage or blastocyst transfer and 7 (0.5%) women had 2 cleavage-stage embryos transferred. A total of 896/1489 (60.2%) had natural cycles, 408/1489 (27.4%) had hormone replacement cycle, 184/1489 (12.4%) had letrozole-induced cycle and 1/1489 (0.1%) had stimulated cycle frozen embryo transfer.\nThe median serum 25(OH)D was 58.7 (25th–75th percentile 45.7–75.0] nmol/l (23.5 ng/ml), where 489/1489 (32.8%) and 1000/1489 (67.2%) women were vitamin D deficient (< 50 nmol/l or < 20 ng/ml) and non-deficient (≥ 50 nmol/l or ≥ 20 ng/ml), respectively. Using the Endocrine Society Guidelines criteria, 489/1489 (32.8%), 627/1489 (42.1%), and 373/1489 (25.1%) women were vitamin D deficient (< 50 nmol/l or < 20 ng/ml), insufficient (≥ 50 nmol/l and < 75 nmol/l or ≥ 20 ng/ml and < 30 ng/ml) and sufficient (≥ 75 nmol/l or ≥ 30 ng/ml), respectively.\nThe patient characteristics in the vitamin D deficient and non-deficient groups are shown in Table  1 . There were no significant differences in the women’s age, duration, type, and cause of infertility and antral follicle count between the two groups. Women who were vitamin D deficient had a significantly higher body mass index compared to the non-deficient group although the difference was small [22.7 (20.6–25.4) vs. 21.9 (20.2–24.0),  P  = < 0.001]. A higher proportion of women in the vitamin D non-deficient group had blastocyst transfer compared to the deficient group. The median (25th–75th percentile) vitamin D level in women with cleavage-stage embryos and blastocysts were 57.5 (42.4–75.7) nmol/l and 59.4 (47.1–74.5) nmol/l, respectively, which was not significantly different ( P  = 0.133).\n Table 1 Patient characteristics Parameters Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) ( n  = 489) Vitamin D non-deficient group  ≥ 50 nmol/l (≥ 20 ng/ml) ( n  = 1000) P  values # Index stimulated IVF cycle Age of women at oocyte retrieval (year) 36 (34–38) 36 (34–38) 0.676 Body mass index (kg/m2) 22.7 (20.6–25.4) 21.9 (20.2–24.0)  < 0.001* Non-smoker 440/487 (90.3) 905/984 (92.0) 0.515 Ethnicity  < 0.001*  Chinese 446 (91.2) 969 (96.9)  Non-Chinese 43 (8.8) 31 (3.1) Duration of infertility (years) 4.0 (3.0–6.0) 4.0 (3.0–6.0) 0.426 Cause of infertility n (%) 0.746  Endometriosis 33 (6.7) 50 (5.0)  Tuboperitoneal factor 58 (11.9) 121 (12.1)  Male factor 171 (35.0) 356 (35.6)  Unexplained 117 (23.9) 240 (24.0)  Mixed factors 110 (22.5) 233 (23.3) Type of infertility 0.201  Primary 357 (73.0) 698 (69.8)  Secondary 132 (27.0) 302 (30.2) Antral follicle count 11 (7–16) 11 (7–16) 0.923 Stimulation regimen 0.370  Long GnRH agonist 5 (1.0) 7 (0.7)  GnRH antagonist 459 (93.9) 925 (92.5)  Progestin primed 25 (5.1) 68 (6.8) Total dose of gonadotrophins (IU) 2700 (2025–3300) 2475 (2100–3150) 0.386 Duration of stimulation (days) 11 (10–12) 11 (10–12) 0.044* Peak estradiol level (pmol/l) 12,750 (8026–21,333) 12,840 (8015–20,357) 0.921 Number of oocytes retrieved 12 (7–17) 11 (7–16) 0.258 Number of oocytes normally fertilized 7 (4–10) 6 (4–10) 0.767 Number of cleavage stage embryos in IVF cycle ( n  = 561) 3 (2–3) 3 (2–3) 0.533 Number of blastocysts in IVF cycle ( n  = 928) 4 (3–6) 4 (3–6) 0.248 Endometrial thickness on trigger day (mm) 11.2 (9.7–12.7) 11.1 (9.5–13.0) 0.877 Stage of embryos in frozen transfer 0.018*  Cleavage stage embryos  Blastocysts 205 (41.9) 284 (58.1) 356 (35.6) 644 (64.4) Blastulation rate (only for women undergoing extended culture) 50% (33.3%–66.7%) ( n  = 284) 50% (33.3%–66.7%) ( n  = 644) 0.764 Number of embryos transferred 1.000 (Fisher’s exact test)  1 203 (99.0) 351 (98.6)  2 2 (1.0) 5 (1.4) Number of blastocysts transferred  1 284 (100.0) 644 (100.0) –- Endometrial preparation 0.138  Natural cycle 310 (63.4) 586 (58.6)  Hormonal replacement cycle 116 (23.7) 292 (29.2)  Letrozole 63 (12.9) 121 (12.1)  Gonadotrophin stimulation 0 (0.0) 1 (0.1) Data shown represent the median (25th–75th percentile) for continuous variables and number (%) for categorical variables * Statistically significant\nPatient characteristics\nCleavage stage embryos\nBlastocysts\n205 (41.9)\n284 (58.1)\n356 (35.6)\n644 (64.4)\nData shown represent the median (25th–75th percentile) for continuous variables and number (%) for categorical variables\n* Statistically significant\nThe main outcomes are shown in Table  2 . When analysing the results based on the threshold in the Endocrine Society guideline of 50 nmol/l (20 ng/ml) for vitamin D deficiency, there were no statistically significant differences in live birth rate, pregnancy rate, ongoing pregnancy rate, and miscarriage rate between the vitamin D deficient and non-deficient groups. The live birth rate, pregnancy rate, ongoing pregnancy rate, and miscarriage rate between the vitamin D deficient and non-deficient groups remained insignificant after adjusting for women’s age at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type, and duration of infertility.\n Table 2 Pregnancy outcomes of frozen embryo transfer in the vitamin D deficient group and non-deficient group Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) Vitamin D non-deficient group  ≥ 50 nmol/l (≥ 20 ng/ml) Odds ratio (95% CI) P  values Adjusted odds ratio (95% CI)# Adjusted  P  values Pregnancy rate 208/489 (42.5%) 468/1000 (46.8%) 0.842 (0.677–1.046) 0.121 0.888 (0.705–1.118) 0.312 Ongoing pregnancy rate 153/489 (31.3%) 361/1000 (36.1%) 0.806 (0.640–1.015) 0.067 0.844 (0.662–1.076) 0.171 Live birth rate 151/489 (30.9%) 341/998 (34.2%) 0.861 (0.683–1.086) 0.205 0.905 (0.710–1.155) 0.423 Miscarriage rate 53/208 (25.5%) 120/468 (25.6%) 0.992 (0.682–1.443) 0.965 0.964(0.657–1.414) 0.852 Ectopic pregnancy 3/208 (1.4%) 2/468 (0.4%) 3.413 (0.566–20.408) 0.173 @ 3.025 (0.483–18.924) 0.237 Stillbirth – 1/468 (0.2%) – 1.000 @ – – Multiple pregnancy 2/208 (1.0%) 6/468 (1.3%) 0.747 (0.150–3.731) 1.000 @ 0.677 (0.134–3.412) 0.636 Data shown represent the median (25th–75th percentile) or number (%) # Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility @ Fisher’s exact test 1 woman in the vitamin D deficient group and 2 women in the vitamin D non-deficient group had termination of pregnancy after ongoing pregnancy. Another 2 women in the vitamin D non-deficient group were loss to follow-up after documented ongoing pregnancy, one of whom was last seen at the antenatal clinic at 28 weeks and subsequently emigrated abroad\nPregnancy outcomes of frozen embryo transfer in the vitamin D deficient group and non-deficient group\nData shown represent the median (25th–75th percentile) or number (%)\n# Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility\n@ Fisher’s exact test\n1 woman in the vitamin D deficient group and 2 women in the vitamin D non-deficient group had termination of pregnancy after ongoing pregnancy. Another 2 women in the vitamin D non-deficient group were loss to follow-up after documented ongoing pregnancy, one of whom was last seen at the antenatal clinic at 28 weeks and subsequently emigrated abroad\nWhen vitamin D insufficiency was analysed among deficient, insufficient, and sufficient groups, there were no statistically significant differences in the live birth rates [151/489 (30.9%), 218/627 (34.8%), 123/371 (33.2%),  P  = 0.391] among the vitamin D deficient, insufficient and sufficient groups, respectively. No statistically significant differences were found in the pregnancy rates, ongoing pregnancy rates, and miscarriage rates among the three groups (Table  3 ).\n Table 3 Pregnancy outcomes of frozen embryo transfer using the Endocrine Society Guidelines cut-off n  (%) P  values @ Adjusted odds ratio (95% CI) # Adjusted  P  values Pregnancy rate 0.348  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 208/489 (42.5%) 0.893 (0.670–1.190) 0.439  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 299/627 (47.7%) 1.009 (0.769–1.323) 0.948  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 169/373 (45.3%) Ref – Ongoing pregnancy rate 0.130  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 153/489 (31.3%) 0.809 (0.599–1.091) 0.165  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 227/627 (36.2%) 0.934 (0.706–1.236) 0.633  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 134/373 (35.9%) Ref – Live birth rate 0.421  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 151/489 (30.9%) 0.906 (0.670–1.227) 0.524  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 218/627 (34.8%) 1.002 (0.754–1.330) 0.991  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 123/371 (33.2%) Ref – Miscarriage rate 0.811  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 53/208 (25.5%) 1.024 (0.635–1.652) 0.922  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 79/299 (26.4%) 1.098 (0.707–1.707) 0.676  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 41/169 (24.3%) Ref – Ectopic pregnancy 0.303  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 3/208 (1.4%) 2.082 (0.206–21.018) 0.534  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 1/299 (0.3%) 0.526 (0.032–8.540) 0.651  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 1/169 (0.6%) Ref – Stillbirth 0.938  Vitamin D deficient group  < 50 nmol/l – – –  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 1/299 (0.3%) – –  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) – – – Multiple pregnancy 0.797  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 2/208 (1.0%) 1.539 (0.138–17.222) 0.726  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 5/299 (1.7%) 3.072 (0.353–26.758) 0.309  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 1/169 (0.6%) Ref – Data shown represent the number (%) @ Chi-square test for trend # Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility\nPregnancy outcomes of frozen embryo transfer using the Endocrine Society Guidelines cut-off\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nData shown represent the number (%)\n@ Chi-square test for trend\n# Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility\nSubgroup analysis on pregnancy outcome was performed for women who had single blastocyst transfer only and cleavage-stage embryos only respectively and no statistically significant differences were found in the pregnancy rates, ongoing pregnancy rates, live birth rates, and miscarriage rates between the vitamin D deficient vs vitamin D non-deficient groups or among the vitamin D deficient, insufficient and sufficient groups (Tables  4 ,  5 ,  6 , and  7 ).\n Table 4 Pregnancy outcomes of frozen embryo transfer in the vitamin D deficient group and non-deficient group (for those with single blastocyst transfer only) Vitamin D deficient group  < 50 nmol/l  (< 20 ng/ml) Vitamin D non-deficient group  ≥ 50 nmol/l  (≥ 20 ng/ml) Odds ratio (95% CI) P  values Adjusted odds ratio (95% CI)# Adjusted  P  values Pregnancy rate 156/284 (54.9%) 369/644 (57.3%) 0.908 (0.686–1.203) 0.502 0.896 (0.675–1.190) 0.449 Ongoing pregnancy rate 117/284 (41.2%) 292/644 (45.3%) 0.845 (0.637–1.121) 0.241 0.835 (0.628–1.112) 0.218 Live birth rate 116/284 (40.8%) 273/642 (42.5%) 0.934 (0.703–1.239) 0.633 0.924 (0.694–1.232) 0.590 Miscarriage rate 36/156 (23.1%) 90/369 (24.4%) 0.930 (0.598–1.447) 0.747 0.947 (0.605–1.483) 0.812 Ectopic pregnancy 3/156 (1.9%) 1/369 (0.3%) 7.194 (0.745–71.429) 0.081 @ 6.707 (0.666–67.539) 0.106 Stillbirth – 1/369 (0.3%) – 1.000 @ – – Multiple pregnancy 2/156 (1.3%) 4/369 (1.1%) 1.185 (0.215–6.536) 1.000 @ 0.961 (0.171–5.398) 0.964 Data shown represent the median (25th–75th percentile) or number (%) # Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, type and duration of infertility @ Fisher’s exact test Table 5 Pregnancy outcomes of frozen embryo transfer using the Endocrine Society Guidelines cut-off (for those with single blastocyst transfer only) n  (%) P  values @ Adjusted odds ratio (95% CI) # Adjusted  P  values Pregnancy rate 0.430  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 156/284 (54.9%) 0.843 (0.590–1.204) 0.348  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 237/418 (56.7%) 0.910 (0.654–1.267) 0.577  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 132/226 (58.4%) Ref – Ongoing pregnancy rate 0.097  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 117/284 (41.2%) 0.723 (0.506–1.032) 0.074  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 182/418 (43.5%) 0.800 (0.576–1.111) 0.183  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 110/226 (48.7%) Ref – Live birth rate 0.406  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 116/284 (40.8%) 0.840 (0.586–1.202) 0.340  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 173/418 (41.4%) 0.863 (0.620–1.202) 0.383  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 100/224 (44.6%) Ref – Miscarriage rate 0.655  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 36/156 (23.1%) 1.168 (0.661–2.066) 0.592  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 63/237 (26.6%) 1.372 (0.819–2.299) 0.229  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 27/132 (20.5%) Ref – Ectopic pregnancy 0.218  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 3/156 (1.9%) 2.055 (0.199–21.174) 0.545  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) – – –  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 1/132 (0.8%) Ref – Stillbirth 0.951  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) – – –  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 1/237 (0.4%) – –  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) – – – Multiple pregnancy 0.338  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 2/156 (1.3%) – –  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)  (≥ 20 ng/ml and < 30 ng/ml) 4/237 (1.7%) – –  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) – – – Data shown represent the number (%) @ Chi-square test for trend # Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility Table 6 Pregnancy outcomes of frozen embryo transfer in the Vitamin D deficient group and non-deficient group (for those with cleavage-stage embryos only) Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) Vitamin D non-deficient group  ≥ 50 nmol/l (≥ 20 ng/ml) Odds ratio (95% CI) P  values Adjusted odds ratio (95% CI)# Adjusted  P  values Pregnancy rate 52/205 (25.4%) 99/356 (27.8%) 0.882 (0.597–1.304) 0.530 0.884 (0.592–1.321) 0.549 Ongoing pregnancy rate 36/205 (17.6%) 69/356 (19.4%) 0.886 (0.568–1.383) 0.594 0.883 (0.557–1.397) 0.594 Live birth rate 35/205 (17.1%) 68/356 (19.1%) 0.872 (0.556–1.366) 0.550 0.875 (0.551–1.390) 0.573 Miscarriage rate 17/52 (32.7%) 30/99 (30.3%) 1.117 (0.543–2.299) 0.763 0.997 (0.471–2.109) 0.994 Ectopic pregnancy – 1/99 (1.0%) – 1.000 @ – – Stillbirth – – – – – – Multiple pregnancy – 2/99 (2.0%) – 0.545 @ – – Data shown represent the median (25th–75th percentile) or number (%) # Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, type and duration of infertility @ Fisher’s exact test Table 7 Pregnancy outcomes of frozen embryo transfer using the Endocrine Society Guidelines cut-off (for those with cleavage-stage embryos only) n  (%) P  values @ Adjusted odds ratio (95% CI) # Adjusted  P  values Pregnancy rate 0.941  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 52/205 (25.4%) 1.020 (0.619–1.682) 0.937  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) 62/209 (29.7%) 1.267 (0.781–2.057) 0.338  Vit D sufficient group (≥ 75 nmol/L)  (≥ 30 ng/ml) 37/147 (25.2%) Ref – Ongoing pregnancy rate 0.875  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 36/205 (17.6%) 1.123 (0.626–2.014) 0.698  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) 45/209 (21.5%) 1.479 (0.843–2.596) 0.173  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 24/147 (16.3%) Ref – Live birth rate 0.851  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 35/205 (17.1%) 1.152 (0.638–2.080) 0.639  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) 45/209 (21.5%) 1.562 (0.885–2.754) 0.124  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 23/147 (15.6%) Ref – Miscarriage rate 0.701  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) 17/52 (32.7%) 0.675 (0.270–1.687) 0.401  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) 16/62 (25.8%) 0.522 (0.211–1.292) 0.160  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 14/37 (37.8%) Ref – Ectopic pregnancy 0.896  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) – – –  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) 1/62 (1.6%) – –  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) – – – Stillbirth –  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) – – –  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) – – –  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) – – – Multiple pregnancy 0.264  Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) – – –  Vit D insufficient group (≥ 50 nmol/L and < 75 nmol/L) (≥ 20 ng/ml and < 30 ng/ml) 1/62 (1.6%) 0.967 (0.039–23.679) 0.983  Vit D sufficient group (≥ 75 nmol/L) (≥ 30 ng/ml) 1/37 (2.7%) Ref – Data shown represent the number (%) @ Chi-square test for trend # Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility\nPregnancy outcomes of frozen embryo transfer in the vitamin D deficient group and non-deficient group (for those with single blastocyst transfer only)\nData shown represent the median (25th–75th percentile) or number (%)\n# Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, type and duration of infertility\n@ Fisher’s exact test\nPregnancy outcomes of frozen embryo transfer using the Endocrine Society Guidelines cut-off (for those with single blastocyst transfer only)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D insufficient group (≥ 50 nmol/L and < 75 nmol/L)\n(≥ 20 ng/ml and < 30 ng/ml)\nData shown represent the number (%)\n@ Chi-square test for trend\n# Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility\nPregnancy outcomes of frozen embryo transfer in the Vitamin D deficient group and non-deficient group (for those with cleavage-stage embryos only)\nData shown represent the median (25th–75th percentile) or number (%)\n# Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, type and duration of infertility\n@ Fisher’s exact test\nPregnancy outcomes of frozen embryo transfer using the Endocrine Society Guidelines cut-off (for those with cleavage-stage embryos only)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVit D sufficient group (≥ 75 nmol/L)\n(≥ 30 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nVitamin D deficient group\n< 50 nmol/l (< 20 ng/ml)\nData shown represent the number (%)\n@ Chi-square test for trend\n# Adjusted for age of women at oocyte retrieval, body mass index, antral follicle count, stage of embryos in frozen transfer, type and duration of infertility\nThe patient characteristics in the vitamin D deficient, insufficient, and replete groups are shown in Table  8 .\n Table 8 Patient characteristics in 3 groups Parameters Vitamin D deficient group  < 50 nmol/l (< 20 ng/ml) ( n  = 489) Vitamin D insufficient group  ≥ 50 nmol/l and < 75 nmol/l (≥ 20 ng/ml and < 30 ng/ml) ( n  = 627) Vitamin D sufficient group  ≥ 75 nmol/l (≥ 30 ng/ml) ( n  = 373) P  values # Index stimulated IVF cycle Age of women at oocyte retrieval (year) 36 (34–38) 36 (34–37) 36 (34–38) 0.705 Body mass index (kg/m2) 22.7 (20.6–25.4) 22.0 (20.2–24.3) 21.9 (20.3–24.3)  < 0.001* Non-smoker 440/487 (90.3) 569/618 (92.1) 336/366 (91.8) 0.820 Ethnicity  < 0.001*  Chinese 446 (91.2) 612 (97.6) 357 (95.7)  Non-Chinese 43 (8.8) 15 (2.4) 16 (4.3) Duration of infertility (years) 4.0 (3.0–6.0) 4.0 (3.0–6.0) 4.0 (3.0–6.0) 0.175 Cause of infertility  n  (%) 0.801  Endometriosis 33 (6.7) 32 (5.1) 18 (4.8)  Tuboperitoneal factor 58 (11.9) 77 (12.3) 44 (11.8)  Male factor 171 (35.0) 217 (34.6) 139 (37.3)  Unexplained 117 (23.9) 160 (25.5) 80 (21.4)  Mixed factors 110 (22.5) 141 (22.5) 92 (24.7) Type of infertility 0.440  Primary 357 (73.0) 437 (69.7) 261 (70.0)  Secondary 132 (27.0) 190 (30.3) 112 (30.0) Antral follicle count 11 (7–16) 11 (7–16) 10 (7–15) 0.355 Stimulation regimen 0.269  Long GnRH agonist 5 (1.0) 5 (0.8) 2 (0.5)  GnRH antagonist 459 (93.9) 573 (91.4) 352 (94.4)  Progestin primed 25 (5.1) 49 (7.8) 19 (5.1) Total dose of gonadotrophins (IU) 2700 (2025–3300) 2475 (2138–3150) 2475 (2100–3150) 0.663 Duration of stimulation (days) 11 (10–12) 11 (10–12) 11 (10–12) 0.041* Peak estradiol level (pmol/l) 12,750 (8026–21,333) 13,255 (8064–21,028) 12,350 (7911–17,968) 0.511 Number of oocytes retrieved 12 (7–17) 12 (7–17) 10 (7–15) 0.019* Number of oocytes normally fertilized 7 (4–10) 7 (4–10) 6 (4–9) 0.185 Number of cleavage stage embryos in IVF cycle ( n  = 561) 3 (2–3) ( n  = 205) 2 (2–3) ( n  = 209) 3 (2–3) ( n  = 147) 0.223 Number of blastocysts in IVF cycle ( n  = 928) 4 (3–6) ( n  = 284) 4 (3–6) ( n  = 418) 4 (3–6) ( n  = 226) 0.495 Endometrial thickness on trigger day (mm) 11.2 (9.7–12.7) 11.1 (9.5–13.0) 11.2 (9.5–13.1) 0.755 Stage of embryos in frozen transfer 0.010*  Cleavage stage embryos 205 (41.9) 209 (33.3) 147 (39.4)  Blastocysts 284 (58.1) 418 (66.7) 226 (60.6) Blastulation rate (only for women undergoing extended culture) 50.0% (33.3%–66.7%) ( n  = 284) 48.5% (33.3%–63.6%) ( n  = 418) 50.0% (33.3%–66.7%) ( n  = 226) 0.252 Number of embryos transferred 0.601  1 203 (99.0) 207 (99.0) 144 (98.0)  2 2 (1.0) 2 (1.0) 3 (2.0) Number of blastocysts transferred –  1 284 (100.0) 418 (100.0) 226 (100.0) Endometrial preparation 0.044*  Natural cycle 310 (63.4) 353 (56.3) 233 (62.5)  Hormonal replacement cycle 116 (23.7) 189 (30.1) 103 (27.6)  Letrozole 63 (12.9) 85 (13.6) 36 (9.7)  Gonadotrophin stimulation 0 (0.0) 0 (0.0) 1 (0.1) Data shown represent the median (25th–75th percentile) for continuous variables and number (%) for categorical variables * Statistically significant # Kruskal–Wallis test for continuous variables\nPatient characteristics in 3 groups\n3 (2–3)\n( n  = 205)\n2 (2–3)\n( n  = 209)\n3 (2–3)\n( n  = 147)\n4 (3–6)\n( n  = 284)\n4 (3–6)\n( n  = 418)\n4 (3–6)\n( n  = 226)\n50.0% (33.3%–66.7%)\n( n  = 284)\n48.5% (33.3%–63.6%)\n( n  = 418)\n50.0% (33.3%–66.7%)\n( n  = 226)\nData shown represent the median (25th–75th percentile) for continuous variables and number (%) for categorical variables\n* Statistically significant\n# Kruskal–Wallis test for continuous variables\n\nOur study showed that serum vitamin D level was not associated with the live birth rate in women undergoing frozen embryo transfer. There were no statistically significant differences in the pregnancy rates, ongoing pregnancy rates, and miscarriage rates between vitamin D deficient and non-deficient groups. However, more women in the vitamin D non-deficient group had blastocyst transfer than cleavage stage embryo transfer compared to the vitamin D deficient group.\nWe have previously published a retrospective study at our unit which found that the cumulative live birth rate of the first IVF cycle, including the fresh and all frozen embryo transfers from one stimulated IVF cycle, in the vitamin D deficient group was significantly lower compared to the non-deficient group [ 15 ]. In that study, women with vitamin D deficiency had a higher body mass index and had less oocytes retrieved and normally fertilized oocytes despite requiring a higher dosage of gonadotrophin for ovarian stimulation compared to the vitamin D non-deficient group, although the magnitude of absolute difference was small. The usual practice in our unit is for extended culture to blastocyst if there are more than 3 cleavage-stage embryos. In our current study on frozen embryo transfer, the dosage of gonadotrophins stimulation, number of oocytes retrieved and number of normally fertilized oocytes in the index IVF cycle were similar in the vitamin D deficient and non-deficient groups. Taking the two studies together, one hypothesis is that vitamin D may act on enhancing oocyte or embryo quality by improving blastocyst formation rather than on endometrial receptivity in affecting pregnancy outcome. However, there were no differences in the blastulation rate between the vitamin D deficient and non-deficient groups when only women with extended culture to blastocysts were considered. We did not find statistically significant differences in endometrial thickness between the vitamin D deficient and non-deficient groups.\nOur findings are in contrast to one study of 368 infertile women undergoing single blastocyst transfer which showed that vitamin D deficiency compromised pregnancy rates by 40%, but similar to 2 recent studies on the association of vitamin D and frozen embryo transfer which did not find serum vitamin D status to be related to pregnancy outcomes [ 21 ,  22 ]. In the study by Franasiak et al., only women with euploid blastocyst transfers were included to control for embryonic factor, although both fresh or frozen embryo transfers were included and the results may not be applicable to women who did not undergo extended embryo culture or preimplantation genetic testing for euploid blastocyst transfer [ 21 ]. A prospective observational cohort study by van de Vijver et al. included 280 women who had 1–2 blastocyst(s) transferred [ 22 ]. In these studies, the rationale of including only blastocyst transfers or even euploid blastocysts was to minimize the biases related to oocyte and embryo quality and actually evaluate whether vitamin D deficiency independently compromises the chances of pregnancy in those reaching the blastocyst stage. However, this would imply the inclusion of a selected ‘better prognosis’ group. We included a larger sample size of women from two tertiary reproductive medicine centres undergoing frozen embryo transfer. Our study population was more heterogenous and included all women who had frozen embryo transfer during the study period, resulting in a mix of women undergoing cleavage stage embryos and blastocyst transfer, but may be more generalizable to the general infertile population. Most (99.5%) of the women in our study had single cleavage-stage embryo or blastocyst transfer. We included the first frozen embryo transfer in our centre as it is generally the practice that better-quality embryos are transferred first. However, we may not be able to entirely exclude all women who may have had embryo transfers or previous IVF cycles prior to the study period.\nExisting studies on vitamin D and frozen embryo transfer involved mostly Caucasian population and serum vitamin D was assessed using immunoassay [ 20 – 22 ]. A retrospective study by Rudick et al. suggested that the relationship between vitamin D status and pregnancy rates differed by race, with vitamin D deficiency leading to lower pregnancy rates in Caucasian women, but the opposite in Asians [ 26 ]. They cautioned that results from their study need to be validated by future cohort studies, given that only 49 out of 188 women were of Asian origin. In another study by the same group on recipients of donor oocytes, which found a lower pregnancy rate in recipients of egg donation who were vitamin D deficient compared to the vitamin D replete group, Asians were the only racial group for which there was no evidence of a beneficial effect of vitamin D albeit the small numbers [ 16 ]. On the contrary, the majority of our patients were Chinese, which may be an important contribution to the controversy about the association of vitamin D and FET outcomes with regard to ethnicity. However, similarly, we were not able to assess for the effect of ethnicity owing to small number of non-Chinese population. Further research is needed to elucidate racial differences in the effect of vitamin D on reproduction. In our study, serum 25(OH)D was assessed by mass spectrometry in an accredited laboratory, which is the gold standard.\nIn our study on frozen embryo transfer, we still found a high prevalence of vitamin D deficiency (32.8%) and insufficiency (42.1%). We adopted the most widely used classification by The Endocrine Society of serum vitamin D replete status as > 75 nmol/l (> 30 ng/ml), deficiency as < 50 nmol/l (< 20 ng/ml) and insufficiency as that in between these levels [ 25 ]. For the purpose of the main analysis, we have grouped the vitamin D insufficient and replete status together as ‘non-deficient’ state. Existing studies and meta-analysis vary in the classification of the intermediate group of ‘insufficient’ vitamin D status, some grouping it together with the replete group as in our study, while others may group it together with the deficient group [ 27 ]. When analysed as three separate groups, we still did not find statistically significant differences in the live birth rates, pregnancy rates, ongoing pregnancy rates and miscarriage rates among the three groups. There is still no consensus on the appropriate cut-off level to define vitamin D deficiency for non-skeletal effects or whether vitamin D supplementation is beneficial in these conditions [ 28 ,  29 ]. Vitamin D levels exist as a continuum and women with vitamin D level at very similar levels close to the cut-off may have been classified into different categories based on very small differences. There is evidence that even higher thresholds of vitamin D beyond 75 nmol/L are beneficial for reproduction in women with polycystic ovary syndrome undergoing ovulation induction [ 30 ]. Our analysis is limited by the small number of women with serum vitamin D beyond these levels.\nThe main limitation of our study is its retrospective and cross-sectional nature. Livebirth was our primary outcome. Indeed, there was a long way to go from frozen embryo transfer until live birth and women may be on various supplements during the course of pregnancy, which we have not documented. We also looked at the ongoing pregnancy rate and early miscarriage rate of our cohort and there were no statistically significant differences in these parameters between the vitamin D deficient and non-deficient groups. A systematic review and meta-analysis have shown an increased risk of miscarriage in naturally conceived women who were vitamin D deficient compared with those who were vitamin D replete [ 31 ]. On the contrary, vitamin D did not seem to be associated with miscarriage in several meta-analyses that included women undergoing IVF [ 10 ,  12 ,  27 ]. A recently published Mendelian randomization study also showed no evidence of a causal relationship between miscarriage and serum 25(OH)D level [ 32 ].\nConcerns have been raised regarding the potential degradation of serum 25(OH)D with storage which may potentially affect interpretation in retrospective studies [ 33 ,  34 ]. Our prospective observational study involving 55 reproductive-aged women undergoing assisted reproductive treatment where we assessed the change in serum 25(OH)D with cryostorage showed that the median percentage decrease in serum 25(OH)D was only 3.4% (statistically not significant) when the serum samples were stored at − 20 °C for 7 months and analysed with mass spectrometry (unpublished). Nevertheless, being a retrospective study based on a convenience sample size, our study may be underpowered to assess the association of vitamin D on pregnancy rate if the effect of vitamin D was too small to be shown in this study. There was a 5% difference in ongoing pregnancy rate between the vitamin D deficient and non-deficient groups although not statistically significant. Our retrospective power analysis suggested that our current sample size would only give a power of 45% to detect a difference of 5%. Although it is arguable whether a difference of 5% is clinically meaningful, Vitamin D is a safe and cheap intervention and vitamin D deficiency can be easily corrected by dietary supplements. This small improvement in pregnancy rate may potentially be cost-effective and relevant to infertile women undergoing in IVF taken into consideration the safety of intervention. Future adequately powered interventional studies on the effect of vitamin D on the live birth rate in infertile women undergoing frozen embryo transfer are needed to answer the question on whether vitamin D supplement should be provided.\n\nSerum vitamin D level was not associated with the live birth rate in women undergoing frozen embryo transfer.","source_license":"CC-BY-4.0","license_restricted":false}