Vitamin D is associated with reduced risk of Sjögren’s syndrome: a Mendelian randomization study

Dear Editor, SS is an autoimmune disease characterized by lymphocytic infiltration of exocrine glands, but also systemic features such as arthritis, vasculitis, pulmonary and neurological dysfunction. Despite its impact on quality of life and physical function, no disease-modifying drugs have been approved. Some observational studies have linked vitamin D deficiency to increased risk of SS [1], but these findings are susceptible to confounding and reverse causation. Clinical trial evidence suggests that vitamin D supplementation reduces overall autoimmune disease risk [2], but whether this applies to SS is not known. Mendelian randomization (MR) uses genetic variants as instrumental variables to estimate the effect of an exposure on an outcome that is typically more robust to these biases. Prior MR studies found no evidence for a link between vitamin D and SS, but were limited by poor case definition and/or statistical power [3, 4]. The aim of this MR study was to estimate the effect of genetically predicted vitamin D level on risk of primary SS.

Genetic data for primary SS was taken from the largest genome-wide association study (GWAS) to date of 6098 cases (defined by American–European Consensus Group criteria) and 34 928 controls [5]. We used data from a GWAS of 25 hydroxyvitamin D (25OHD) concentration in 417 580 individuals of European ancestry [6]; associations were adjusted for age, sex, assessment month, assessment centre, first 40 principal components, genotyping batch, and supplement intake. To minimize risk of collider bias, primary analysis used genetic associations unadjusted for BMI. Because some lead variants for 25OHD are associated with body composition, we also used data that conditioned on BMI as a sensitivity analysis. To minimize potential bias due to horizontal pleiotropy, we selected variants (P < 5 × 10⁻⁸, r² < 0.1 and present in the SS GWAS) from four gene regions previously shown to be strongly associated with 25(OH)D and implicated in the transport, metabolism, and synthesis of vitamin D—GC, DHCR7, CYP2R1 and CYP24A1 [7]. Our instrument selection differed from prior one-sample (applying step-wise conditional selection to individual-level data [3, 7]) and two-sample MR analyses (using variants across the genome [4]). We used the inverse-variance weighted method and pleiotropy robust sensitivity analyses, i.e. MR Egger and weighted median/mode methods. Inverse-variance weighted and Egger methods additionally accounted for weak correlation between variants. MR estimates were scaled to per S.D. (∼21 nmol/l) increase in 25OHD. This analysis used publicly available summary statistics that had already obtained ethical approval.

Fifteen variants were selected to instrument 25OHD (Supplementary Table S1, available at Rheumatology online) that together explained 3.2% of the variance in 25OHD concentration, corresponding to a mean F statistic of 931. Genetically predicted 25OHD was associated with reduced risk of SS (odds ratio 0.72 per S.D. decrease in 25OHD; 95% CI 0.57, 0.90; P = 0.004). Estimates were directionally concordant in pleiotropy robust sensitivity analyses and analysis accounting for BMI (Supplementary Fig. S1) and appear to be mostly driven by rs11023374 in the CYP2R1 gene (Supplementary Fig. S2).

This study is the first to provide genetic evidence that higher circulating vitamin D levels may reduce the risk of primary SS. Our analyses differ from earlier MR studies of vitamin D that were limited by the small number of poorly defined cases, which may explain their null findings [3, 4]. MR studies have failed to replicate most traditional observational associations between vitamin D and autoimmune diseases (with the exception of MS [8]), likely due to confounding and/or reverse causation in traditional designs. The VITAL trial supported a role of vitamin D for reducing autoimmune diseases but was not powered to interrogate individual diseases [2]. The shared immune mechanisms (abnormal B and T cell response to autoantigens) between MS and SS may provide mechanistic insight into the potential effect of vitamin D on autoimmunity [3].

One limitation is that risk factors for disease incidence, which were the focus of this study, may not be relevant to progression of established disease. The key assumption of no pleiotropic pathways cannot be empirically verified. However, we tried to mitigate this by selecting instruments from genes relevant to vitamin D biology. The sensitivity analyses were also consistent with the main results. In summary, we found genetic evidence to support a causal link between vitamin D and SS. Vitamin D supplementation could represent a relatively safe and inexpensive therapeutic option, but these findings need to be confirmed in randomized controlled trials. These results have implications for potential disease prevention strategies and the interpretation and design of future trials.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

Summary statistics for SS is accessible in the Databases of Genotypes and Phenotypes (dbGaP) under accession number phs002723.v1.p1. Summary statistics for vitamin D can be accessed via https://cnsgenomics.com/content/data.

Contribution statement

S.S.Z. had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Funding

S.S.Z. is supported by a National Institute for Health Research Clinical Lectureship and works in centres supported by Vs Arthritis (grant no. 21173, 21754 and 21755). SB is supported by the Wellcome Trust (225790/Z/22/Z) the United Kingdom Research and Innovation Medical Research Council (MC_UU_00002/7) and the National Institute for Health Research Cambridge Biomedical Research Centre (NIHR203312). The views expressed are those of the authors and not necessarily those of the National Health Service, the National Institute for Health Research, or the Department of Health and Social Care. The funder had no role in the design or conduct of the study; the collection, management, analysis and interpretation of the data; the preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Disclosure statement: The authors have declared no conflicts of interest.