Polycystic ovary syndrome (PCOS) is one of the most common endocrine disorders affecting 6–10% of reproductive age women [
Until now, metformin has been the first-line treatment for patients with PCOS who have insulin resistance, but associated side effects such as abdominal pain, diarrhea, or headache are common [
Vitamin D deficiency is defined as a 25-hydroxyvitamin (25-OH)D serum concentration of less than 20 ng/ml, while vitamin D insufficiency occurs at a level from 20 ng/ml to <30 ng/ml [
The protocol for this systematic review was registered in the International Prospective Register of Systematic Reviews (Prospero) as number CRD42020157444 [
Pubmed, Embase, Web of Science, Cochrane Library, and Clinical Trials were searched up to March 22, 2020, with the following MeSH and non-MeSH terms: “vitamin D”, “vitamin D3”, “vitamin D2”, “cholecalciferol”, “ergocalciferols”, “hydroxycholecalciferol” or “calcitriol” combined with “polycystic ovary syndrome”, “ovary polycystic disease”, “PCOS”, “stein-Leventhal syndrome”, “stein-Leventhal syndrome”. The reference lists of identified literatures were also browsed to identify any potential additional publications. No restrictions were made for language or date of publications.
Inclusion criteria were as follows: (1) RCTs; (2) study subjects were women diagnosed with PCOS; (3) PCOS was diagnosed on the basis of the 2003 Rotterdam criteria or the 1990 National Institute of Child Health and Human Development criteria; (4) full text was accessible; (5) studies comparing the therapeutic effect of vitamin D supplement with placebo.
Exclusion criteria were as follows: (1) studies examining the effects of vitamin D combination with other interventions such as metformin, calcium, oral contraceptive, and so on; (2) studies examining the effects of vitamin D among patients with PCOS undergoing intrauterine insemination (IUI) or in vitro fertilization (IVF) treatment; (3) incomplete data; (4) genetic research.
Literature searches were conducted by two reviewers (S. G. and H. Y. J.) separately, and then the title and abstract were screened for eligibility. Full texts retrieved were carefully checked according to the inclusion and exclusion criteria to select qualified trials independently. Disputes were solved through discussion with another reviewer (T. Y.), and consensus was reached.
Data extraction were carried out by two reviewers (S. G. and H. Y. J.), and the following information was extracted: the last name of the first author, publishing year, country, criteria used for diagnosis of PCOS, study population, sample size, type and duration of intervention, dose of vitamin D intake, serum vitamin D level, biochemical indices of glucose and lipid metabolism including fasting plasma glucose (FPG), fasting insulin, HOMA-IR, QUICKI, serum triglyceride (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), very low-density lipoprotein cholesterol (VLDL-C), hypersensitive C-reactive protein (hs-CRP), total testosterone (Total T), sex hormone-binding globulin (SHBG), and dehydroepiandrosterone sulfate (DHEAS). Any discrepancies were resolved by discussion with another reviewer (T. Y.). In the event of incomplete information, authors were contacted to acquire relevant data.
The Cochrane risk of bias assessment tool was employed to evaluate the method of randomization, allocation of concealment, blinding of participants, personnel and outcome assessment, incomplete outcome, selective reporting, and other biases. Risk bias of each study was graded as low, unclear, and high. Any dispute was resolved by consensus.
Statistical synthesis and subgroup analysis were performed by Revman V.5.3 offered by the Cochrane Collaboration. Measurement data were displayed as mean difference and standard deviation. Mean difference (MD) with 95% confidence intervals (CI) by inverse variance method was employed in case of data with identical measuring units; otherwise, standard mean difference (SMD) was adopted. Heterogeneity among studies was estimated by Cochran’s Q test and I-squared.
The online database search yielded 329 articles. After removing 147 duplicates, 182 articles were excluded by screening the title and abstract, and then, the remaining 61 articles were inspected carefully for eligibility. Studies without outcomes of interest (
Flowchart of search strategy.
These papers were published from 2012 to 2019 and conducted in Iran [
Summary of RCTs focusing on the effect of Vitamin D supplementation in women with polycystic ovary syndrome.
Author/year/country | Population/age/vitamin D level | Intervention | Duration | 25OHD before treatment(g/ml) | 25OHD after treatment (ng/ml) | Hyperandrogenism | Insulin resistance | Dyslipidemia | Inflammation |
---|---|---|---|---|---|---|---|---|---|
Seyyed et al., (2017) | 36 women with PCOS, | G1: 50,000 IU/week of oral vitamin D3 ( | 8 weeks | G1: 3.5 ± 4.2; | 28.24 ± 6.47; | ↓FPG ↔HOMA-IR, | |||
Asemi et al., (2017) | 104 overweight and obese women with PCOS, | G1: 1000 mg/d calcium + vitamin D placebo ( | 8 weeks | G1: 13.9 ± 2.0; | 71.2 ± 14.7; | ↓insulin, HOMA-IR, | ↓TG, VLDL-C | 8 weeks | |
Bonakdaran et al., (2012) | 48 women with PCOS, | G1: 1000 mg/d metformin ( | 12 weeks | G1:28.2 ± 13.5; | 26.7 ± 10.6; | ↔Total testosterone, | ↔FPG, insulin, | ||
Foroozanfard et al., (2015) | 104 overweight women with PCOS who have vitamin D deficiency, | G1: 1000 mg calcium/d plus vitamin D placebo weekly ( | 8 weeks | G1: +0.3 ± 0.4; | ↓hs-CRP | ||||
Gupta et al., (2017) | 50 women with PCOS, | G1: 12000 IU/week vitamin D ( | 12 weeks | G1: 18.56 ± 9.68; | 44.90 ± 9.04; | ↔-Total testosterone, | ↓FPG, | ↔TG, TC, | |
Rahimi-Ardabili et al., (2013) | 50 women with PCOS, | G1: 50,000 IU/20 days of oral cholecalciferol ( | 8 weeks | G1:6.9 ± 2.8; | 23.4 ± 6.14; | ↓ TC | ↓hs-CRP | ||
Javed et al. (2019) | 37 women with PCOS, | G1: 3200IU/day vitamin D ( | 12 weeks | G1:10.26 ± 4.57; | 36.22 ± 7.81; | ↔Total testosterone, | ↔FPG, insulin, | ↔TG, TC, LDL-C, | ↔ hs-CRP |
Irani et al., (2015) | 53 women with PCOS, | G1: 50,000 IU/week of oral vitamin D3 ( | 8 weeks | G1: 16.3 ± 0.9; | 43.2 ± 2.4; | ↔Total testosterone, | ↔ HOMA-IR | ↓TG | |
Maktabi et al., (2017) | 60 women with PCOS, | G1: 50,000 IU/2 weeks of oral vitamin D3 ( | 12 weeks | G1: 12.8 ± 4.5; | 27.5 ± 9.8; | ↔Total testosterone, | ↓FPG, insulin, | ↔TG, TC, LDL-C, | ↓hs-CRP |
Jorly Mejia-Montilla et al., (2018) | 169 women with PCOS, | G1: 5000 IU/day vitamin D ( | 12 weeks | G1:13.7 ± 4.2; | 19.1 ± 4.9; | ↓FPG, insulin, | ↓TG, TC, LDL-C | ||
Ardabili et al., (2012) | 50 women with PCOS, | G1:50,000 IU of oral vitamin D3 every 20 days ( | 8 weeks | G1: 6.9 ± 2.8; | 23.4 ± 6.1; | ↔FPG, insulin, | |||
Jafari-Sfidvajani et al., (2017) | 56 women with PCOS, | G1: low-calorie diet+ 50,000 IU/week oral vitamin D3 ( | 12 weeks | G1: 15.83 ± 4.85; | 43.38 ± 12.61; | ↔Total testosterone, | |||
Trummer et al., (2018) | 123 women with PCOS, | G1:20,000 IU of oral vitamin D3/week ( | 24 weeks | G1: 19.55 ± 6.73; | 36.14 ± 8.05; | ↔Total testosterone | ↔ FPG, HOMA-IR, | ↔TG, TC |
IU: international units; FPG: fasting plasma glucose; IR: insulin resistance; HOMA-IR: homeostasis model of assessment-estimated insulin resistance; QUICKI: quantitative insulin sensitivity check index; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; VLDL-C: very low-density lipoprotein cholesterol; TC: total cholesterol; TG: triglyceride; hs-CRP: high sensitive C-reactive protein; SHBG: sex hormone-binding globulin; DHEAS: dehydroepiandrosterone sulfate.
Quality assessment for randomized controlled trials included on basis of Cochrane risk of bias assessment tool.
Twelve studies reported significant increase in serum vitamin D level following supplementation. The baseline serum level of 25(OH)D in intervention group ranged from 3.5 ± 4.2 ng/ml to 19.55 ± 6.73 ng/ml, indicating that selected patients with PCOS were mostly vitamin D deficient. Combination of data extracted from thirteen studies revealed a significant increase in 25(OH)D concentrations in the vitamin D treatment group versus placebo (MD: 16.19 ng/ml, 95% CI: 13.30, 19.09) (Figure
Forest plot of vitamin D concentration in experimental and placebo groups.
Nine studies were evaluated for FPG level. Meta-analysis showed serum FPG level of patients with PCOS significantly decreased after vitamin D supplementation (SMD: −0.34, 95% CI: −0.61, −0.07) (Figure
(a) Forest plot of FPG level in experimental and placebo groups. (b) Forest plot of FPG level in women with PCOS supplemented by day, week, and 20 days. (c) Forest plot of FPG level in women with PCOS who have vitamin D deficiency or insufficiency. (d) Forest plot of FPG level in women with PCOS supplemented with low- or high-dose vitamin D.
In total, ten studies have assessed insulin resistance indices in women with PCOS, among which seven studies reported concentration of fasting insulin, and ten studies reported HOMA-IR and six studies evaluated QUICKI. We found a significant decrease in serum fasting insulin level (SMD: −0.43, 95% CI: −0.67, −0.18) (Figure
(a) Forest plot of fasting insulin level in experimental and placebo groups. (b) Forest plot of HOMA-IR in experimental and placebo groups. (c) Forest plot of QUICKI in experimental and placebo groups.
A total of eight trials were included for assessing the effect of vitamin D supplementation on lipids, all of which reported serum triglyceride levels. Among these studies, seven trials reported serum concentrations of total-C and HDL-C, six trials reported results for LDL-C, and three trials examined serum level of VLDL-C. The overall effect of vitamin D supplementation significantly lowered serum VLDL-C level (MD: −3.83 mg/dl, 95% CI: −7.34, 0.32; Figure
(a) Forest plot of total-C level in experimental and placebo groups. (b) Forest plot of LDL-C level in experimental and placebo groups. (c) Forest plot of VLDL-C level in experimental and placebo groups. (d) Forest plot of HDL-C level in experimental and placebo groups. (e) Forest plot of triglycerides in experimental and placebo groups. (f) Forest Plot of triglycerides in women with PCOS who have vitamin D deficiency or insufficiency. (g) Forest plot of total-C in women with PCOS who have vitamin D deficiency or insufficiency. (h) Forest plot of triglycerides in women PCOS supplemented by day, week, two weeks, and 20 days. (i) Forest plot of total-C in women with PCOS supplemented by day, week, and 20 days. (j) Forest plot of triglycerides in women with PCOS supplemented with a low or high dose. (k) Forest plot of total-C in women with PCOS supplemented with a low dose or high dose. (l) Forest plot of LDL-C in women with PCOS supplemented with a low dose or high dose.
Seven studies in total reported on serum androgens, of which five studies had results for total testosterone concentration, five trials had data on DHEAS, and four trials examined serum SHBG level. The results of the meta-analysis revealed that vitamin D supplementation in women with PCOS did not have significant effect on total testosterone (SMD: −0.18, 95% CI: −0.37, 0.02; Figure
(a) Forest plot of total testosterone level in experimental and placebo groups. (b) Forest plot of DHEAS level in experimental and placebo groups. (c) Forest plot of SHBG level in experimental and placebo groups. (d) Forest plot of hs-CRP level in experimental and placebo groups.
The funnel plots for serum vitamin D concentrations indicated a risk for lack of reporting on negative effect of vitamin D supplementation. The funnel plots for the rest of the indices showed there was no significant publication bias. However, for each result, the number of studies of meta-analysis was less than 10, which may be too small to determine publication bias through funnel plots (Figures
Funnel plot of standard error by standard differences in the means of plasma: (a) vitamin D; (b) FPG; (c) fasting insulin; (d) HOMA-IR; (e) QUICKI; (f) triglycerides; (g) total-C; (h) HDL-C; (i) LDL-C; (j) VLDL-C; (k) total testosterone; (l) DHEAS; (m) SHBG; (n) hs-CRP in selected trials.
Several meta-analyses have been conducted concerning the effect of vitamin D supplementation on metabolic biomarkers of women with PCOS, suggesting variable beneficial effects, but the results remained conflicting. Moreover, the data of non-RCTs and cosupplementation trials were combined in most of these reviews. In view of the increasing number of RCTs regarding this topic, we conducted the present meta-analysis of RCTs focusing on the effect of vitamin D intake alone without cosupplementation compared with placebo to reach more convincing conclusions. The results of our meta-analysis revealed that vitamin D oral intake alone improved insulin resistance parameters and reduced inflammation in patients with PCOS. Furthermore, subgroup analysis showed that lipid metabolism was also improved in vitamin D deficient group. No effects were found on serum androgen levels or inflammation status.
Vitamin D deficiency is very prevalent in women with PCOS. A recent cross-sectional study demonstrated that, compared with fertile controls, significantly lower vitamin D levels were present in women with PCOS (mean 25(OH)D of 64.5 nmol/l vs. 49.0 nmol/l, resp.); meanwhile, higher HOMA-IR and lipid abnormalities are associated with deficient vitamin D levels [
Evidence from our meta-analysis showed that vitamin D supplementation resulted in lowering blood fasting glucose levels in addition to improving insulin resistance, as seen by a significant decrease in serum fasting insulin and HOMA-IR along with a slight increase in QUICKI. The results of our meta-analysis are in contrast with earlier ones by other researchers. Xue et al. [
Notably, decline in serum triglycerides, total-C, and VLDL-C concentration were significant among patients with PCOS who are deficient in vitamin D in our meta-analysis. This finding is in accordance with the significant improvement of serum triglycerides level observed in the meta-analysis by Xue et al. [
In this meta-analysis, we did not find any improvement in serum total testosterone, DHEAS, or SHBG level. Similar findings were reported in the meta-analysis by Azadi-Yazdi et al. [
Our meta-analysis has several strengths. It includes comprehensive research of RCTs and strict inclusion of high-quality studies, in which vitamin D intervention was given alone. This allowed for a more focused analysis of vitamin D effects in PCOS. Our work also has some limitations. There is considerable heterogeneity between studies with variable race/ethnicity and age. In addition, baseline vitamin D status, vitamin D dose, and formulation and duration of treatment of vitamin D vary between studies and are an additional source of confounding. Moreover, most of the trials included were conducted in Iran, where women are covered by clothes resulting in less exposure to sunlight, so whether the benefits of vitamin D supplementation can be generalizable to women with PCOS from all over the world is still an open question. In addition, considering the limited number of high-quality RCTs and small sample size, the conclusions of the present meta-analysis should be extrapolated with caution. However, as an inexpensive and safe treatment option, vitamin D supplementation can be implemented in women with PCOS who have vitamin D deficiency. Our study may shed light on the potential mechanisms behind those discovered benefits that vitamin D deficiency may be a codeterminant factor in metabolism disorder. Taken together, vitamin D supplementation appears to influence several aspects of clinical features present in women with PCOS (Figure
Potential mechanism of therapeutic effect of vitamin D on women with PCOS.
Our systematic review showed that oral intake of vitamin D supplementation attenuated insulin resistance and hyperlipemia, but not androgenic profile or inflammatory markers in women with PCOS who have vitamin D deficiency. Several potential mechanisms may be underlying these beneficial effects of vitamin D on clinical features of PCOS, and further research is needed to explore the complex role of vitamin D in different pathways of this metabolic disorder.
The data used in the study are available upon request to the corresponding author.
The authors report no conflicts of interest related to this work.
The authors acknowledge the instruction of statistical methods by Yang Wang. This work was supported by the National Natural Science Foundation of China (81471520), Beijing Natural Science Foundation Project (7182054), and Beijing Nova Programme Interdisciplinary Cooperation Project (Z191100001119015).
Supplementary Figure S1: (a) forest plot of fasting insulin level in women with PCOS who have vitamin D deficiency or insufficiency; (b) forest plot of HOMA-IR in women with PCOS who have vitamin D deficiency or insufficiency; (c) forest plot of QUICKI in women with PCOS who have vitamin D deficiency or insufficiency; (d) forest plot of HOMA-IR in women with PCOS supplemented by day, week, two weeks, and 20 days; (e) forest plot of fasting insulin in women with PCOS supplemented by day, week, two weeks, and 20 days; (f) forest plot of QUICKI in women with PCOS supplemented by week, two weeks, and 20 days; (g) forest plot of fasting insulin level in women with PCOS supplemented with low- or high-dose vitamin D; (h) forest Plot of QUICKI in women with PCOS supplemented with low- or high-dose vitamin D; (i) forest plot of HOMA-IR in women with PCOS supplemented with low- or high-dose vitamin D.