Vitiligo is an acquired pigmented skin disease caused by the loss of melanocyte function, which is characterized by localized or generalized complete depigmentation of skin mucosa. It can occur in all parts of the body, commonly in finger back, wrist, forearm, face, neck and around genitalia, affecting both men and women (1). Globally, the prevalence of vitiligo was estimated to be ranging from a low of 0.5% to a high of 2.0% in adults, and the peak onset period is between 10 and 30 years of age (2). Although vitiligo does not bring fatal risk, it is often accompanied by a variety of autoimmune system-related diseases, such as Hashimoto's thyroiditis, diabetes, Addison's disease and so on (3–6). Therefore, it is necessary to further explore the influence of modifiable risk factors on the incidence of vitiligo, which may provide new ideas for its prevention and treatment.

Previous studies have identified a series of nutritional factors associated with vitiligo, such as selenium (7), copper, and zinc (8). Recently, vitamin D deficiency has been reported to be a potential risk factor for vitiligo (9). With usual physiological intake, the majority of vitamin D is converted to 25-hydroxyvitamin D (25(OH)D) and released into the blood, thus circulating 25(OH)D concentrations can reflect vitamin D status in the human body. As the main metabolite of 25(OH)D, the levels of circulating 25-hydroxy vitamin D₃ [25(OH)D₃] is also widely accepted as an indicator of overall vitamin D status (10). Though no randomized controlled trials have been conducted to investigate the preventive role of vitamin D on the risk of vitiligo to date (11), numerous observational epidemiological studies have evaluated circulating 25(OH)D and 25(OH)D₃ levels in patients with vitiligo relative to controls. For instance, Atazadeh et al. (12) conducted a case-control study with 90 vitiligo patients and 90 healthy controls in Iran, and found that the median (interquartile range) of circulating 25(OH)D level was lower in patients with vitiligo than in matched healthy controls [28.50 (22.52–33.20) ng/mL vs. 29.10 (27.40–35.70) ng/mL, P = 0.01). Another case-control study (13) conducted in India with 100 vitiligo patients and 100 healthy controls suggested that circulating 25(OH)D₃ levels were lower in vitiligo patients than controls (16.17 ± 8.63 vs. 25.49 ± 1.02 ng/ml, P < 0.001). However, Saniee (14) recruited 98 patients with vitiligo and 98 matched healthy controls and measured 25(OH)D levels, but did not detect statistically significant differences between patients with vitiligo and normal subjects (22.37 ± 10.78 vs. 26.16 ± 17.11 ng/ml, P = 0.19). Similarly, Ustun et al. also found no statistically significant association (15). These inconsistent results may be due to differences in the study population, sample size, measurement methods and so on. Besides, findings from traditional observational studies are susceptible to bias such as confounding and reverse causation, therefore, it remains unclear whether the observed association was causal or not.

Mendelian Randomization (MR) study utilizes genetic variants such as single nucleotide polymorphisms (SNPs), which act as instrumental variables (IVs) to estimate the potential causal relationship between exposure and outcome (16). Since genotypes are randomly distributed in the process of gamete formation, the relationship between exposure and outcome of interest will not be biased by common confounding factors, such as postnatal environment, socio-economic status and behavioral factors in conventional observational studies. Moreover, MR study has unique advantages in causal inference because genetic variation is inherited from parents and remains unchanged after birth, so the association between genetic variation and outcome is reasonable in time sequence.

Therefore, in the present study, we first performed a systematic review and meta-analysis of existing observational evidence on the association of circulating 25(OH)D or 25(OH)D₃ levels with risk of vitiligo, and then performed MR analysis to further investigate whether there was evidence for a causal relationship of circulating 25(OH)D and 25(OH)D₃ levels with risk of vitiligo.

To date, vitiligo is a disease that cannot be completely cured. Patients often encounter psychological difficulties due to discrimination, such as shame, depression and anxiety, which usually lead to inferiority and social isolation (47, 48), and have an adverse impact on the quality of life (49). At present, the pathogenic mechanism of vitiligo is not completely clear. There are studies suggesting that vitamin D may increase the melanogenesis and tyrosinase content of human melanocytes through its anti-apoptotic effect, thus preventing the loss of skin pigment (50). In addition, vitamin D can also inhibit the autoimmune pathway in the pathogenesis of vitiligo by inhibiting the expression of interleukin-6 (IL-6), interleukin-8 (IL-8), tumor necrosis factor α (TNF-α) and tumor necrosis factor γ (TNF-γ), regulating the mature differentiation and activation of dendritic cells and inhibiting antigen presentation (51). For prevention and early diagnosis of vitiligo, the present study investigated the role of circulating 25(OH)D and 25(OH)D₃ on the risk of vitiligo by using meta-analysis and Mendelian randomization, however, we did not find convincing evidence to support a protective role of vitamin D supplementation on the risk of vitiligo.

The results from the meta-analysis indicated that patients with vitiligo had a lower level of 25(OH)D compared with controls. Consistently, a meta-analysis conducted by Zhang et al. (52) in 2018 including 679 patients and 537 controls reported similar results (SMD = −0.94, 95% CI: −1.39, −0.48, P = 0.0001). Our updated meta-analysis included five additional recent published studies with a sample size of 987 patients and 770 controls for 25(OH)D, 305 patients and 321 controls for 25(OH)D₃, and therefore had increased statistical power. Similarly, we also found heterogeneity in the studies included in the meta-analysis. However, meta-regression did not suggest potential sources of heterogeneity, and subgroup analysis consistently revealed associations between circulating levels of 25(OH)D and risk of vitiligo in different strata. Publication bias was detected for the association between 25(OH)D and risk of vitiligo, suggesting that our results may be more based on positive results, resulting in overestimation of the summary effect. Previously, no meta-analysis has focused on the association between 25(OH)D₃ and risk of vitiligo. In the present study, we found no statistical differences in circulating levels of 25(OH)D₃ between cases and controls. However, the number of included studies was relatively small, more studies are warranted for obtaining stable results. Additionally, owing to the limitations inherent in observational studies, high heterogeneity among studies may be due to differences in the study design, characteristics of the study population such as age and ethnicity, use of drugs and so on. Finally, all of the included studies were cross-sectional studies and case-control studies with relatively small study sample sizes, which is weak in demonstrating causality, since it is difficult to judge the time sequence of exposure and disease (53). Therefore, low vitamin D levels could also be a consequence of already developed vitiligo.

In order to overcome the limitations inherent in conventional observational studies, we adopted a two-sample MR method which used genetic variants as IVs of exposures to further investigate the relationship of 25(OH)D and 25(OH)D₃ levels with risk of vitiligo. MR approach offers great opportunities to the etiological research of diseases, provided that the following three assumptions are satisfied (54). The first assumption is that the IVs must be associated with the exposure of interest. To ensure this, we used independent loci associated with circulating 25(OH)D and 25(OH)D₃ levels achieving genome-wide significance level (P < 5 × 10⁻⁸) as IVs, which was identified from the largest GWAS to date. The second assumption is that the IVs must not be associated with potential confounders of the exposure-outcome association. Since genotypes are randomly allocated during gamete formation, MR analyses using genetic variants as IVs to a large extent solve the problem of confounding in conventional observational studies. The third assumption is that the IVs influence the outcome only through the risk factor. To ensure this, we excluded potential pleiotropic SNPs and retained those solely associated with 25(OH)D and 25(OH)D₃ levels in the MR sensitivity analyses to evaluate the robustness of our results. We also performed different MR methods to test for potential pleiotropy. We did not observe evidence of directional pleiotropy for the causal association between 25(OH)D, 25(OH)D₃ and risk of vitiligo in any of the above MR approaches. However, the estimate of MR studies reflect the effects of lifelong interference, whereas that of observational studies reflects more acute effects.

The limitations of the present study should be noted. First, the data for the meta-analysis were from observational studies, in which there could be potential confounding factors in the baseline characteristics of the selected population. The statistical ability of the small sample is limited, thus the well-designed and large sample studies are necessary to validate the findings. Second, heterogeneity in meta-analysis is a potential problem that affects the results of statistical analysis. In the present study, subgroup analysis and meta-regression did not indicate sources account for the heterogeneity observed in the study. In addition, publication bias was found in the results of 25(OH)D, which indicated that the results of meta-analysis were affected by publication bias. Third, the variance explained by IVs is limited, with 3.95% total 25(OH)D and 4.58% 25(OH)D₃, respectively. Besides, the number of reported cases in the outcome dataset is relatively small, which could affect the statistical power in MR study. However, it is the largest sample size in the outcome database available at present, which needs more vitiligo GWASs with larger sample size to explore the association. Finally, because our MR analyses were restricted to participants of European ancestry, it is unclear whether our findings can be extrapolated to other study populations. Therefore, further prospective studies and functional studies in vivo and in vitro are needed to elucidate the exact role of 25(OH)D and 25(OH)D₃ in the occurrence of vitiligo.

Our study showed that although there seemed to be an inverse association between vitiligo and 25(OH)D level based on meta-analysis of observational studies, MR analysis pointed to a lack of causal association. In addition, meta-analysis and MR study did not provide evidence of an associations between 25(OH)D₃ level and risk of vitiligo. The findings suggested there is no convincing evidence that vitamin D may help to prevent vitiligo.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

JS performed the literature review, conducted data analysis, interpreted findings, and drafted the manuscript. KL and WC carried on data analysis and interpreted findings of meta-analysis. BL and HY mainly conducted on the data collation of the Mendelian randomization study. JS, LL, and XS can take responsibility for statistical reports, tables, and figures of the data analysis. YM and DY directed analytic strategy, supervised the study from conception to completion and revised drafts of the manuscript. All authors have read and approved the final manuscript.

This work was supported by grants from the National Natural Science Foundation of China (81973663), the Natural Science Foundation of Zhejiang Province (LQ21H260001 and LQ20H260008), the Talent Project of Zhejiang Association for Science and Technology (2018YCGC003), the Foundation of Zhejiang Chinese Medical University (2020ZG01 and 2020ZG16), the Research Project of Zhejiang Chinese Medical University (2021JKZKTS004A).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.