Atrial fibrillation (AF) is the most common persistent cardiac rhythm disorder encountered in clinical practice, and is associated with increased risks of stroke, dementia, heart failure (HF), and death (1). Currently, the estimated prevalence of AF in adults is between 2 and 4% and is strongly associated with advancing age, placing extensive economic and societal burden in terms of morbidity, mortality, increased healthcare resource utilization and loss of productivity (2, 3).

Vitamin D is an essential nutrient obtained from exposure to sunlight, diet, and dietary supplements (4). In the skin, solar ultraviolet B radiation converts 7-dehydrocholesterol to previtamin D₃, which is immediately isomerized to vitamin D₃. This precursor form of vitamin D is metabolized to 25-hydroxyvitamin D [25(OH)D] in the liver, which is used to determine vitamin D status in the routine clinical setting. 25-hydroxyvitamin D is biologically inactive and must be converted to its active form 1,25-dihydroxyvitamin D by 25-hydroxyvitamin D-1-α-hydroxylase mainly in the kidneys (4). Vitamin D deficiency is a global problem, as many as 37% of the general population worldwide having vitamin D deficiency under the definition of <20 ng/mL (<50 nmol/L). Using a definition of <12 ng/mL (<30 nmol/L), approximately 7% of the population may be at risk to manifest symptoms due to severe vitamin D deficiency (5).

In some observational studies, a low vitamin D level has been linked with a variety of cardiovascular diseases such as coronary artery disease (CAD), HF and AF (6, 7). However, these associations remain controversial. Moreover, given the nature of observational studies, these results may be confounded by multiple variables. Mendelian randomization (MR) uses genetic variants as instrumental variables to make causal inferences, avoiding potential biases caused from confounding and reverse causation in observational studies. Therefore, a MR study can be considered as analogous to a randomized controlled trial (RCT), inferring a causal association of the risk factor on the disease outcome (8). Whilst current evidence shows that there may be a relationship between vitamin D deficiency and AF, incomparable study designs and methodological limitations make existing evidence less reliable. In this study, we aimed to perform a two-sample MR analysis to determine the potential role of circulating vitamin D status on AF risk.

The characteristics of 6 SNPs and their association with 25(OH)D and AF are presented in Table 1. None of the individual 6 SNPs was associated with AF at Bonferroni corrected significance level P < 0.008 [(P < 0.05)/6 SNPs], and the allele frequencies between the exposure and outcome population were generally similar (Table 1). The F-statistic of each SNPs was more than 10, which indicated that they were all strong instrumental variables. There was no sample overlap between the exposure and outcome population (Table 1).

In the overall analysis based on IVW method, the OR of AF per unit increase in log 25(OH)D was 1.003 (95% CI, 0.841–1.196; P = 0.976) for random-effects method and 1.003 (95% CI, 0.876–1.148; P = 0.968) for fixed-effect method, which was consistent with the results from weighted median [OR, 0.960, 95% CI (0.826–1.117); P = 0.600], MR-Egger [OR, 0.924, 95% CI (0.653–1.307); P = 0.678], simple median [OR, 0.984, 95% CI (0.776–1.247); P = 0.894] and penalized weighted median method [OR, 0.960, 95% CI (0.825–1.118); P = 0.602], indicating that the circulating 25(OH)D concentrations were not significantly associated with AF (Figure 2). Scatterplot depicting the relationship of the SNP effects on the 25(OH)D against the SNP effects on AF is shown in Supplementary Figure 1.

Heterogeneity among the causal estimates of the six SNPs was not observed (Q = 8.476, P = 0.132). There was no evidence of pleiotropy using MR-Egger (intercept = 0.004, se = 0.008, P = 0.610) or MR-PRESSO global test (P = 0.312). No outlier SNPs were identified in the MR-PRESSO analysis. According to leave-one-out analysis (Supplementary Figure 2) and single SNP analysis (Figure 3), the association between circulating 25(OH)D and AF was not driven by any individual SNP.

In the present MR analysis, we have capitalized on the summary statistics of two large meta-GWAS conducted for serum 25(OH)D and AF in European populations, without any sample overlap, and constructed a strong instrumental variable for 25(OH)D from six SNPs. We have employed a range of MR methods to investigate the association between 25(OH)D and AF. However, none of these analyses suggested a causal relationship between circulating 25(OH)D concentrations and AF in the general population.

As an essential nutrient for individuals, vitamin D performs an important role in the regulation of calcium, phosphorus, and bone metabolism. Besides, vitamin D receptors (VDR) are found in a variety of cells and tissues, including endothelial cells, vascular smooth muscle cells, cardiac myocytes and fibroblasts (25). Thus, it exerts many other important cellular regulatory functions include modulation of oxidative stress status, inflammatory response, mitochondrial function, insulin, and renin-angiotensin–aldosterone system (RAAS) (25).

In experimental study, Hanafy et al. (26) found that vitamin D increased the action potential duration and contractility in isolated rabbit left atrium tissue, revealed a direct electromechanical effect of vitamin D on preventing or terminating AF. Moreover, in our previous meta-analysis investigating the relationship between circulating vitamin D levels and AF risk which included observational studies mainly assessing for chronic AF, the results suggested that vitamin D deficiency modestly increases the risk of AF (7). More recently, another meta-analysis of observational studies also suggested that serum vitamin D deficiency (<20 ng/ml) was associated with an increased risks of AF in the general population and postoperative AF in patients underwent coronary artery bypass graft (CABG) (27). However, in the VITAL study (28), a large randomized clinical trial that recruited 25,119 adults for a median 5.3 years of daily doses (2,000 IU) of vitamin D supplementation and follow-up, the results did not support the use of vitamin D for the primary prevention of new-onset AF. Therefore, data concerning the causal association between vitamin D and AF are so far contradictory and controversial.

Currently, there are various studies looking at cardiovascular effects of vitamin D using MR, designed to minimize bias from confounding, however, most of these studies have found null effects (29). For example, a one-sample MR in 92,416 individuals of Danish descent failed to demonstrate any evidence of a causal association between plasma 25(OH)D levels and CAD or myocardial infarction (30). Similarly, a 2020 study in 417,580 Europeans from the UK Biobank found no evidence to suggest that vitamin D concentration is associated with CAD (31). In regarding to hypertension, the evidence from existing MR studies on the effects of predicted serum 25OHD levels on hypertension, systolic and diastolic blood pressure, consistently does not support any of these outcomes (29). In the current study, we also found no evidence that genetically predicted 25(OH)D is associated with increased risk of AF in the general population. Results of these MR studies suggests that the observed association of reduced 25(OH)D with CAD and AF may not represent a causal relationship, but more likely is due to confounding by unaccounted influences or due to reverse causation.

Notably, a recent non-linear MR study have suggested an L-shaped association of 25(OH)D with CVD risk, with increased CVD risk largely restricted to individuals with low vitamin D status (32). Moreover, in another similar study (33), the authors have demonstrated non-linear dose-response relationships of 25(OH)D with CAD, stroke and all-cause mortality in observational analysis, and identified an association of genetically-predicted 25(OH)D with all-cause mortality only for individuals with vitamin D deficiency (<25 ng/mL) but not for general population. Therefore, further MR studies using non-linear design and focusing on the vitamin D deficiency population are needed. Besides, categorical analysis using different cutpoints proposed for vitamin D deficiency is also warranted in order to identify the subsets of individuals who may benefit more from vitamin D supplementation. Additionally, in a MR study conducted in a small sized Chinese cohort of cardiac outpatients with stable CAD, in contrast to our study that evaluated individuals primarily form European ancestry, the result suggested genetically deprived vitamin D exposure is associated with increased risk of AF (34). Thus, the null effects observed in our study may also be due to the racial disparity. Future studies focusing the effects of vitamin D on risk of AF among different ethnic groups are needed.

Genetically predicted circulating vitamin D concentrations, especially for a non-deficient range, were not causally associated with AF in the general population. Future studies using non-linear design and focusing on the vitamin D deficiency population are needed to further evaluate the causal effect of 25(OH)D concentrations on AF.

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

The ethical reviews and informed consent had been obtained in the original studies, and were not required for the current analysis.

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

This work was supported by grants from the National Natural Science Foundation of China (Grant Number: 82100341 to ZZ, 81970270 and 82170327 to TL).

GYHL has been a consultant and speaker for BMS/Pfizer, Boehringer Ingelheim and Daiichi-Sankyo. No fees are received personally. The remaining 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.