Table 1.
Observational studies.
| Study | Country | Population | Exclusion Criteria | Age, Years (Range, Mean or Median) | Study Design | AMH Assay | Relationship between Vitamin D and AMH | Covariates Adjusted | Vitamin D Status |
|---|---|---|---|---|---|---|---|---|---|
| Healthy/Normal Menstrual Cycle | |||||||||
| Merhi 2012 [42] |
USA | Premenopausal women (n = 388) with regular menstrual cycles who were either HIV positive or high-risk HIV negative | Cancer; hepatic disease; prior hysterectomy/oophorectomy; pregnant | 25–45 | Cross-sectional | DSL ELISA | Serum 25OH-D was positively correlated with serum AMH levels in late-reproductive-age women (≥40 years old); a weak negative correlation between serum vitamin D and AMH was noted in young individuals (<35 years of age) |
HIV status BMI Race Smoking Current drug use Fasting Glucose Insulin level EGFR Geographic site |
14 (9–21) ng/mL |
| Chang 2014 [43] |
South Korea | 73 healthy women | Obesity; history of infertility; systemic disease; smoking; medication or nutritional supplements in last 3 months; irregular menstrual cycles | 27–38 | Cross-sectional | Gen II ELISA | 25(OH)D did not correlate with AMH or other ovarian reserve markers | None | Baseline 25(OH)D concentration of the study population (in winter): 10.3 +/− 4.6 ng/mL |
| Fabris 2017 [44] |
Spain | Healthy oocyte donors (n = 851) with regular menses and at least six antral follicles per ovary divided into 3 groups according to vitamin D level (<20, 20–30, >30 ng/mL) | PCOS women | 18–35 | Cross-sectional | Elecsys automated assay (Roche) | No correlation between serum AMH and bioavailable vitamin D; no correlation between AFC and bioavailable vitamin D | Age BMI |
29.5% were vitamin D replete (>30 ng/mL); 52% had vitamin D deficiency (20–30 ng/mL), and 18.5% had insufficient vitamin D (<20 ng/mL) |
| Kim 2018 [45] |
South Korea | 291 premenopausal women with regular period | Hysterectomy and ovarian surgery; chemotherapy or radiation; estrogen suppressive therapy; OCPs; medications; calcium or vitamin D supplements; abnormal thyroid function test; undetectable AMH |
35–49 | Cross-sectional | AMH Gen II | There was no correlation between AMH and 25OH-D after adjustment for age. | Age | 76.6%, of subjects were serum vitamin D-insufficient (<20 ng/mL); mean vitamin D level 15.9 ng/mL |
| Jukic 2018 [46] |
USA | 825 women aged 30 to 44 years without any known fertility problems | Hx of infertility PCOS Endometriosis Partner with infertility Recurrent breastfeeding |
30–44 | Cross-sectional | ELISA (ANSH Labs) | 25(OH)D was not correlated with AMH, FSH, or inhibin-B. Multivariable results with continuous hormonal outcomes were also null. For dichotomous outcomes, there was a tendency for insufficient 25(OH)D (<30 ng/mL) to be associated with low AMH (<0.7 ng/mL) | Age, Race, BMI, Smoking Hx, Recent use of birth control | 36 ± 11 ng/mL |
| Purdue-Smith 2018 [47] |
USA | US registered nurses who participated in the NHS2 prospective study (1989): women who experienced menopause between time of blood collection and age 45 (cases, n = 328), women who experienced menopause after age 48 (controls, n = 328) | Cancer; Cardiovascular disease; hysterectomy or oophorectomy; radiation or chemotherapy-induced menopause; menopause prior to blood draw | 32–54 | Cross-sectional | Pico AMH assay (ANSH Labs) | Adjusted geometric means of AMH concentrations did not vary according to free 25(OH)D concentration quartiles or total 25(OH)D concentration quartiles | Age; Smoking; BMI; Parity; Physical activity; Breastfeeding; Timing of blood collection; Alcohol intake; Dietary intake; |
Quartiles for total 25(OH)D concentrations (nmol/L) and number of cases:controls Q1, median 44.2, 86:83 Q2, median 59.8, 92:83 Q3, median 71.1, 71:82 Q4, median 90.4, 79:80 |
| Xu 2019 [48] |
China | 33 women with POI, no iatrogenic cause or chromosomal abnormality, and no hormone therapy for at least 6 months; 72 healthy women with regular menstrual cycles and no history of infertility |
Taking vitamin D supplements or other medications that affect vitamin D and ovarian reserve determinants; hysterectomy; oophorectomy; ovarian surgery; chemotherapy or radiation; cigarette smoking; autoimmune disease | 18–40 | Cross-sectional | Electrochem-iluminescence immunoassay (Cobas e602) | 25(OH)D levels were positively but insignificantly correlated with log-transformed AMH, even after adjusting for confounders | Age BMI Education Annual household income |
POI women: 25OH-D: 92.38 +/− 31.07 nmol/L Control women 25OH-D: 96.76 +/− 33.12 nmol/L |
| Women with polycystic ovary syndrome | |||||||||
| Pearce 2015 [49] |
Australia | PCOS (n = 58) and non-PCOS (n = 282) women presenting to fertility clinic | Undetectable serum AMH levels (<3 pmol/L); Consumption of supplements containing more than 500 IU of vitamin D | <40 | Cross-sectional | ELISA (Immunotech) |
Seasonal variations in serum vitamin D were observed between summer and winter, but no seasonal variation in serum AMH levels; no correlation between serum AMH or AFC and vitamin D levels over the year even after adjustment for known confounders. When the cohort was divided into PCOS and ovulatory groups, still no significant relationship was observed. | Age, BMI, Skin color, Menstrual cycle length, Occupation | Summer month 83.4 +/− 5.9 nmol/L Winter month 49.3 +/− 3.6 nmol/L |
| Wong 2018 [50] |
Hong Kong | 451 PCOS women and 244 healthy ovulatory women | Active endocrine or metabolic disease; using any prescription drug; pregnant in the past 3 months |
18–40 | Cross-sectional | Chem-iluminescent immunoassays | Both serum 25(OH)D and AMH levels peaked during summer in PCOS women. In ovulatory women, only serum 25(OH)D but not AMH level showed such seasonal variation. Serum 25(OH)D level in PCOS women significantly correlated positively with AMH, AMH/AFC ratio, and other metabolic parameters; 25(OH)D level was an independent predictor of serum AMH level in women with PCOS but not in ovulatory women. | Age BMI Free androgen index |
74.9% Vitamin D deficient (<20 ng/mL); 21.7% vitamin D insufficient (between 20–29 ng/mL) 3.3% vitamin D sufficient (>30 ng/mL) |
| Bakeer 2018 [51] |
Egypt | 53 PCOS females with infertility and 17 healthy controls | Cushing syndrome, androgen-secreting tumors, congenital adrenal hyperplasia and hyperprolactinemia | 17–39 | Cross-sectional | ELISA or colorimetric | No significant correlation existed between AMH and 25(OH)D, BMI, and dyslipidemia markers. | Age | PCOS 31.32 ± 14.85 (nmol/L) Control 48.65 ± 27.30 (nmol/L) |
| Szafarowska 2019 [52] |
Poland | 25 patients with PCOS and 23 control women | Women on oral hormonal contraception; hormonal intrauterine device | 25–43 | Cross-sectional | DRG ELISA EIA-5738 |
There was no correlation between AMH and 25(OH)D levels in the PCOS or in the control group. Genetic analysis revealed associations between VDR polymorphisms and AMH levels in PCOS women. | Age | Vitamin D levels in PCOS group (14.2 ng/mL) were lower than control group (19.6 ng/mL) |
| Arslan 2019 [53] |
Turkey | 146 infertile women divided into normal ovarian reserve (n = 86) vs. high ovarian reserve (PCOS, n = 60). Women were further divided based on VDD: (Group A) severe VDD (<10 ng/mL, n = 101) and (Group B) mild VDD (10–20 ng/mL, n = 45) | Smoking; prior hysterectomy and/or oophorectomy; endometriosis; ovarian masses; menopause; pregnancy; endocrine disorders; renal dysfunction; hypertension | 18–35 | Cross-sectional | ECLIA method using an automated analyzer (Cobas 6000) | Serum AMH levels were not correlated with 25(OH)D levels in the normal ovarian reserve or PCOS group | Age BMI FSH LH Steroid hormones |
Normal ovarian reserve group 25(OH)D 9.0 ± 6.0 (ng/mL) PCOS group 25(OH)D) 8.5 ± 6.7 (ng/mL |
| Infertile/IVF women | |||||||||
| Neville 2016 [54] |
Ireland | Couples using their own gametes for a fresh IVF/ICSI cycle (n = 64 women) | None | 36.5 ± 3.3 | Cross-sectional | Not reported | No significant correlation between 25(OH)D and AMH | None | Mean serum 25(OH)D concentration 47.4 ± 2.8 nmol/L; 12 deficient (<30 nmol/L), 28 suboptimal (30–50), 24 sufficient (>50) |
| Drakopoulos 2017 [55] |
Belgium | Healthy infertile women (n = 283) undergoing their first infertility treatment divided into vitamin D deficient (<20 ng/mL or normal vitamin D levels (≥20 ng/mL) |
Vitamin D supplementation; medication for systemic disease; iatrogenic (ovarian sx., gonadotoxic therapy) or genetic cause of ovarian reserve loss | 18–42 | Cross-sectional | Gen II ELISA | The mean AMH and AFC levels did not differ significantly between the vitamin D-deficient and the vitamin D-normal groups; No correlation was observed between 25-OH Vitamin D and AMH or AFC | Age BMI Infertility cause Smoking Season |
30.7% (n = 87) had vitamin D < 20 ng/mL; 69.3% (n = 196) had vitamin D >20 ng/mL |
| Lata 2017 [56] |
India | Infertile women with unexplained infertility (n = 35) and fertile controls (n = 35). Both groups were vitamin D-deficient | History of smoking; OCPs; any hormonal or steroid drug use; known VDD, obesity (BMI > 35); endometriosis; thyroid disorders; autoimmune disease; tubal factor, male factor, or PCOS | 18–40 | Cross-sectional | ELISA (ANSH labs) | No correlation between AMH and Vitamin D was found in either group (no values reported). | Age, duration of married life, duration and type of infertility, obstetrical history, education level |
Case 6.18 ± 2.09 ng/mL Control 4.85 ± 3.02 ng/mL |
| Shapiro 2018 [57] |
USA | 457 infertile women with high prevalence of diminished ovarian reserve | All women who had baseline measurements of 25OH-D, AMH, and FSH within 90 days of each other were included |
21–50 | Cross-sectional | Not reported | AMH and FSH levels did not vary between women with VDD and those with normal levels; Multivariate linear regression analysis of log-transformed AMH and FSH with 25OH-D levels adjusted for confounders confirmed lack of association. | Age BMI Seasonal variations |
16.2% (n = 74) had 25OH-D <20.0 ng/mL; 83.8% (n = 383) had 25OH-D ≥20 ng/mL |
| Bednarska-Czerwińska 2019 [58] |
Poland | 53 infertile women (diagnosed with tubal factor infertility and qualified for IVF) with AMH >0.7 ng/mL | Hypertension; diabetes; renal dysfunction; hyperinsulinism; PCOS; endometriosis | 34.7 ± 4.1 | Cross-sectional | ECLIA immunoanalyzer (Cobas e411) | Overall, a nonsignificant negative linear correlation between serum AMH and total vitamin D; However, a change-point was noted; Negative linear correlation between levels of serum AMH and total vitamin D concentrations up to approximately 31 ng/mL; Beyond that threshold, a nonsignificant positive correlation was observed. | Age BMI |
Total vitamin D (ng/mL) in serum overall: 29.7 ± 13.3; During winter/spring: 26.3 ± 13.2; During summer/autumn: 34.2 ± 12.6 |
| Liu 2019 [59] |
China | 848 infertility patients undergoing IVF | Patients with premature ovarian insufficiency; patients treated with ICSI; women whose 25(OH)D levels were taken 4 weeks prior to IVF cycle | 31.67 | Cross-sectional | Not reported | Serum Vitamin D levels were inversely related to AMH, although this was not statistically significant. | None | Patients divided into 4 groups based on serum 25(OH)D quartiles (ng/mL); Group 1: 9.04; Group 2: 13.67; Group 3: 16.20; Group 4: 23.22 |
AFC, antral follicle count; AMH, anti-Müllerian hormone; AUC, area under the curve; BMI, body mass index; ELISA, enzyme-linked immunosorbent assay; EGFR, estimated glomerular filtration rate; FSH, follicle-stimulating hormone; Hx, history; LH, Luteinizing hormone; ICSI, intracytoplasmic sperm injection; 25OH-D, 25-hydroxy vitamin D; OCP, oral contraceptive pills; PCOS, polycystic ovary syndrome; POI, premature ovarian insufficiency; ROC, receiver-operating characteristic curve; VDD, vitamin D deficiency; NHS2, Nurses’ Health Study II; and IVF, in vitro fertilization.